Re: fakeroot status

[Roland McGrath]

> Ha, that was way too easy with rpctrace.  It turns out that /dev/fd/3 is
> looked up in the namespace of the fakeroot translator, which conveniently,
> but in a fundamentally broken way looks up the node and does all the MAGIC
> interpretation itself.

Duh.  One of these days I will switch my brain on ahead of time.

> Now, this starts to look as if using file_name_lookup_under was not the best
> choice.  Can't we use dir_lookup directly?

Yes, we should.  It is hairy, we have to duplicate much of the contents of
hurdlookup.c and lookup-retry.c, that's why file_name_lookup_under was the
easy choice.  This is difficult to do properly.  In fact, it's impossible
with the netfs_attempt_lookup interface.  We have to implement
netfs_S_dir_lookup ourselves, so that it can return partial results using
FS_RETRY_MAGICAL to the user.  It can be a bit simpler than the original
function since we don't need the passive translator stuff, and it would be
fine to deoptimize relative symlinks (i.e. symlinks always return to the user).
You also don't need to do all the iouser diddling or reauth stuff,
since fakeroot trusts all user ports the same anyhow.