linux-kernel - Re: "fs/namei.c: keep track of nd->root refcount status" causes boot panic

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20190903134832.GH1131@ZenIV.linux.org.uk>
Date:   Tue, 3 Sep 2019 14:48:32 +0100
From:   Al Viro <viro@...iv.linux.org.uk>
To:     Christoph Hellwig <hch@...radead.org>
Cc:     Qian Cai <cai@....pw>, linux-fsdevel@...r.kernel.org,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: "fs/namei.c: keep track of nd->root refcount status" causes boot
 panic

On Tue, Sep 03, 2019 at 06:04:56AM -0700, Christoph Hellwig wrote:
> On Tue, Sep 03, 2019 at 01:37:19PM +0100, Al Viro wrote:
> > On Tue, Sep 03, 2019 at 12:21:36AM -0400, Qian Cai wrote:
> > > The linux-next commit "fs/namei.c: keep track of nd->root refcount status” [1] causes boot panic on all
> > > architectures here on today’s linux-next (0902). Reverted it will fix the issue.
> > 
> > <swearing>
> > 
> > OK, I see what's going on.  Incremental to be folded in:
> > 
> > diff --git a/include/linux/namei.h b/include/linux/namei.h
> > index 20ce2f917ef4..2ed0942a67f8 100644
> > --- a/include/linux/namei.h
> > +++ b/include/linux/namei.h
> > @@ -20,8 +20,8 @@ enum {LAST_NORM, LAST_ROOT, LAST_DOT, LAST_DOTDOT, LAST_BIND};
> >  #define LOOKUP_FOLLOW		0x0001	/* follow links at the end */
> >  #define LOOKUP_DIRECTORY	0x0002	/* require a directory */
> >  #define LOOKUP_AUTOMOUNT	0x0004  /* force terminal automount */
> > -#define LOOKUP_EMPTY		0x4000	/* accept empty path [user_... only] */
> > -#define LOOKUP_DOWN		0x8000	/* follow mounts in the starting point */
> > +#define LOOKUP_EMPTY		0x8000	/* accept empty path [user_... only] */
> > +#define LOOKUP_DOWN		0x10000	/* follow mounts in the starting point */
> >  
> >  #define LOOKUP_REVAL		0x0020	/* tell ->d_revalidate() to trust no cache */
> >  #define LOOKUP_RCU		0x0040	/* RCU pathwalk mode; semi-internal */
> 
> Any chance to keep these ordered numerically to avoid someone else
> introdcing this kind of bug again later on?

Not sure what would be the best way to do it...  I don't mind breaking
the out-of-tree modules, whatever their license is; what I would rather
avoid is _quiet_ breaking of such.