Xref: utzoo comp.unix.questions:18428 comp.unix.wizards:19761
Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!wuarchive!udel!haven!adm!smoke!gwyn
From: gwyn@smoke.BRL.MIL (Doug Gwyn)
Newsgroups: comp.unix.questions,comp.unix.wizards
Subject: Re: stat() call on a mounted directory point
Keywords: stat,mount,directory
Message-ID: <11812@smoke.BRL.MIL>
Date: 14 Dec 89 08:06:38 GMT
References: <1279@nih-csl.UUCP>
Reply-To: gwyn@brl.arpa (Doug Gwyn)
Followup-To: comp.unix.questions
Organization: Ballistic Research Lab (BRL), APG, MD.
Lines: 16

In article <1279@nih-csl.UUCP> ted@nih-csl.UUCP (ted persky) writes:
>When I issue a stat() call on any directory which happens to be a mount
>point for a partition (as indicated by 'df'), the inode number that comes
>back is '2', which is identical to the inode of the root directory.

That's what you're supposed to see; it's part of the semantics of mounting.

>Issuing an 'ls -il /' gives you the actual inode number of /usr, which is
>what I want stat() to do.

That's too bad.  "ls" merely reports the inumber stored in its directory
entry, and that is not altered when a filesystem is mounted on the inode.

>P.S. E-mail would be appreciated.  This is kind of urgent.

It can't be TOO urgent since UNIX has always worked like that.