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  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]
Date:   Thu, 9 Aug 2018 20:56:26 -0400
From:   "J. Bruce Fields" <bfields@...ldses.org>
To:     NeilBrown <neilb@...e.com>
Cc:     Jeff Layton <jlayton@...nel.org>,
        Alexander Viro <viro@...iv.linux.org.uk>,
        Martin Wilck <mwilck@...e.de>, linux-fsdevel@...r.kernel.org,
        Frank Filz <ffilzlnx@...dspring.com>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH 3/5] fs/locks: change all *_conflict() functions to
 return a new enum.

On Fri, Aug 10, 2018 at 09:40:35AM +1000, NeilBrown wrote:
> caller_fl is first and sys_fl is second.
> 
> if sys_fl, the second, is a read lock, and caller_fl, the first, is a
> write lock, they clearly conflict but any other lock that conflict
> with caller_fl (The write lock) would *not* necessarily conflict with
> the read lock.  So this situation is *not*  FL_TRANSITIVE_CONFLICT.
> 
> locks_conflict() only returns FL_TRANSITIVE_CONFLICT when sys_fl (the
> second) is a write lock, which it isn't in this case.  So I think that
> this case is handled correctly.
> posix_locks_conflict() will return FL_CONFLICT, but not
> FL_TRANSITIVE_CONFLICT.
> 
> Have I convinced you, or have I missed your point?

Eh, I was just confused.

And now I'm tempted to blame you for confusing me, but maybe that's just
my ego going defensive.

(My bruised ego suggests leaving locks_conflict and its callers alone,
and having an entirely separate function that checks this when we need
it.)

--b.

Powered by blists - more mailing lists