[RFC][PATCH] Pid namespaces vs locks interaction

[Serge E. Hallyn]

Quoting Vitaliy Gusev (vgusev at openvz.org):
> Hello
> 
> On 6 December 2007 18:51:30 Serge E. Hallyn wrote:
> > > fl_pid is used by nfs, fuse and gfs2. For instance nfs  keeps in fl_pid
> > > some unique id to identify locking process between hosts - it is not a
> > > process pid.
> >
> > Ok, but so the struct user_flock->fl_pid is being set to the task's
> > virtual pid, while the struct kernel_flock->fl_pid is being set to
> > task->tgid for nfsd use.
> >
> > Why can't nfs just generate a uniqueid from the struct pid when it
> > needs it?
> 
> I think it is hard. lockd uses struct nlm_host to get process unique id (see 
> __nlm_alloc_pid() function).

Looks pretty simple though...  That whole set of code could even stay
the same except for in __nlm_alloc_pid():

	option 1: compare struct pid* instead of uint32_t pid
	option 2: use the "global pid" out of the stored struct pid,
		something like pid->numbers[0].nr.

> > Fuse just seems to copy the pid to report it to userspace, so it would
> > just copy pid_vnr(kernel_flock->pid) into user_flock->fl_pid.
> >
> > Anyway I haven't looked at all the uses of struct fl_pid, but you
> > can always get the pidnr back from the struct pid if needed so there
> > should be no problem.
> >
> > The split definately seems worthwhile to me, so that
> > user_flock->fl_pidnr can always be said to be the pid in the acting
> > process' namespace, and flock->fl_pid can always be a struct pid,
> > rather than having fl_pid sometimes be current->tgid, or sometimes
> > pid_vnr(flock->fl_nspid)...
> >
> > -serge
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
> > the body of a message to majordomo at vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> 
> 
> -- 
> Thank,
> Vitaliy Gusev