[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: chown32() weirdness

To: Christoph Hellwig <hch@caldera.de>
Subject: Re: chown32() weirdness
From: Stephen Lord <lord@sgi.com>
Date: Mon, 18 Feb 2002 14:58:23 -0600
Cc: Eric Sandeen <sandeen@sgi.com>, Wessel Dankers <wsl@fruit.eu.org>, linux-xfs@oss.sgi.com
References: <Pine.LNX.4.33.0202181038540.17806-100000@chuckle.americas.sgi.com> <Pine.LNX.4.33.0202181050220.17806-100000@chuckle.americas.sgi.com> <20020218181215.A26602@caldera.de>
Sender: owner-linux-xfs@oss.sgi.com
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.7) Gecko/20011226

Christoph Hellwig wrote:

>On Mon, Feb 18, 2002 at 11:02:51AM -0600, Eric Sandeen wrote:
>
>>Ok, I guess this is a difference between Irix and Linux...
>>
>
>There are two possible chown semantics.  Linux has a restricted chown
>by default that only allows root to change owners and thus defines
>_POSIX_CHOWN_RESTRICTED. 
>
>Some other systems (dunno about IRIX) allow both variants, depending
>on configuration.  Then pathconf() with _PC_CHOWN_RESTRICTED is used to
>find out which semantics apply to a specific filename.
>
>	Christoph
>

Irix does support both, and by removing the double permissions check I 
made xfs
fall back to the default Irix behavior. Looks like Eric has around found 
the switch
inside xfs to make this work the linux way.

Steve

References:
- Re: chown32() weirdness
  - From: Eric Sandeen <sandeen@sgi.com>
- Re: chown32() weirdness
  - From: Eric Sandeen <sandeen@sgi.com>
- Re: chown32() weirdness
  - From: Christoph Hellwig <hch@caldera.de>

Prev by Date: Re: xfsdump & compression? (was: Re: xfsdump aborts after assertion failure)
Next by Date: Re: xfsdump & compression? (was: Re: xfsdump aborts after assertion failure)
Previous by thread: Re: chown32() weirdness
Next by thread: Re: chown32() weirdness
Index(es):
- Date
- Thread