Subject: Re: [Regression] 2.6.38 ncpfs
From: "Dr. Bernd Feige" <bernd.feige@uniklinik-freiburg.de>
To: Bongani Hlope <bonganih@bankservafrica.com>
Cc: torvalds@linux-foundation.org, arnd@arndb.de, npiggin@kernel.dk,
        greg@kroah.com, hch@lst.de, bonganilinux@mweb.co.za, rjw@sisk.pl,
        petr@vandrovec.name, linux-fsdevel@vger.kernel.org,
        linux-kernel@vger.kernel.org
In-Reply-To: <4DA815B6020000AC0004595E@CLU_MAIL.BANKSERV.CO.ZA>
References: <4DA815B6020000AC0004595E@CLU_MAIL.BANKSERV.CO.ZA>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 8BIT
Date: Fri, 15 Apr 2011 10:53:50 +0200
Message-ID: <1302857630.2798.0.camel@pst15.ukl.uni-freiburg.de>
Mime-Version: 1.0
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1011
Lines: 26

Hi Bongani, Linus et al.

> > You might also try to replace it with
> > 
> >     BUG_ON(!mutex_is_locked(&dentry->d_parent->d_inode->i_mutex))
> > 
> > ie add that "d_parent" there. Just for testing - I think the real fix
> > really is to remove it, but I'd personally be happier knowing that
> > _if_ it were to have that d_parent there, it would have worked.
> 
> I've made that change and ncpfs works fine with it. I've attached a
> patch that changes the BUG_ON to look at the parent.

Seconded, checked it with the Novell servers in our clinic and it works
like a charm - even with the large directories for which I still have to
revert commit 3825bdb7ed920845961f32f364454bee5f469abb from 2.6.37.6
(https://lkml.org/lkml/2011/1/26/302). Thanks a lot!

Bernd


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/