Re: [PATCH review 52/85] sunrpc: Properly encode kuids and kgids in auth.unix.gid rpc pipe upcalls.

From: Eric W. Biederman
Date: Fri Feb 15 2013 - 13:47:42 EST

"J. Bruce Fields" <bfields@xxxxxxxxxxxx> writes:

> On Wed, Feb 13, 2013 at 03:22:32PM -0800, Eric W. Biederman wrote:
> I've committed his patches to
> git:// for-3.9
> which should also get pulled into next. Let me know if there's another
> patch I should take.

I don't currently see any problems.

My changes are now in:

git:// for-next

So we will get whatever linux-next testing there is.

>> I am a touch concerned about /proc/fs/nfsd/exports after my patches
>> and Stanislavs patches both come in. As I think that will allow for
>> cases where net->user_ns != &init_userns. But we can cross that bridge
>> when we come to it.
> OK, let me know of anything I can do to make your life easier....

/proc/fs/nfsd/exports is read-only so it turns out not be be a problem.

That we can not mount nfs or the nfsd control filesystem outside of
init_userns prevents any other problems. Details are in my conversation
with Stanislav.


p.s. It would be nice if /proc/fs/nfsd/exports became a symlink to
/proc/net/nfsd/exports so the code didn't have to use current to figure
out how to do the right thing.

p.p.s Full user namespace support probably isn't much more than using
net->user_ns and adding the FS_USERNS_MOUNT flag.
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at
Please read the FAQ at