linux-kernel - Re: Inode Lock Scalability V7 (was V6)

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: <20101022023444.GA6573@amd>
Date:	Fri, 22 Oct 2010 13:34:44 +1100
From:	Nick Piggin <npiggin@...nel.dk>
To:	Al Viro <viro@...IV.linux.org.uk>
Cc:	Nick Piggin <npiggin@...nel.dk>,
	Dave Chinner <david@...morbit.com>,
	linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: Inode Lock Scalability V7 (was V6)

On Fri, Oct 22, 2010 at 03:20:10AM +0100, Al Viro wrote:
> On Fri, Oct 22, 2010 at 11:45:40AM +1100, Nick Piggin wrote:
> 
> > No you didn't make these points to me over the past couple of weeks.
> > Specifically, do you agree or disagree about these points:
> > - introducing new concurrency situations from not having a single lock
> >   for an inode's icache state is a negative?
> 
> I disagree.
> 
> > And I have kept saying I would welcome your idea to reduce i_lock width
> > in a small incremental patch. I still haven't figured out quite what
> > is so important that can't be achieved in simpler ways (like rcu, or
> > using a seperate inode lock).
> 
> No, it's not a small incremental.  It's your locking order being wrong;
> the natural one is
> 	[hash, wb, sb] > ->i_lock > [lru]
> and that's one hell of a difference compared to what you are doing.

There is no reason it can't be moved to that lock order (or have
new concurrency situations), but the point is that in the first
lock breaking pass, it does not.


> Look:
> 	* iput_final() should happen under ->i_lock
> 	* if it leaves the inode alive, that's it; we can put it on LRU list
> since lru lock nests inside ->i_lock
> 	* if it decides to kill the inode, it sets I_FREEING or I_WILL_FREE
> before dropping ->i_lock.  Once that's done, the inode is ours and nobody
> will pick it through the lists.  We can release ->i_lock and then do what's
> needed.  Safely.
> 	* accesses of ->i_state are under ->i_lock, including the switchover
> from I_WILL_FREE to I_FREEING
> 	* walkers of the sb, wb and hash lists can grab ->i_lock at will;
> it nests inside their locks.

What about if it is going on or off multiple data structures while
the inode is live, like inode_lock can protect today. Such as putting
it on the hash and sb list.


> 	* prune_icache() grabs lru lock, then trylocks ->i_lock on the
> first element.  If trylock fails, we just give inode another spin through
> the list by moving it to the tail; if it doesn't, we are holding ->i_lock
> and can proceed safely.
> 
> What you seem to miss is that there are very few places accessing inode through
> the lists (i.e. via pointers that do not contribute to refcount) and absolute
> majority already checks for I_FREEING/I_WILL_FREE, refusing to pick such
> inodes.  It's not an accidental subtle property of the code, it's bloody
> fundamental.

I didn't miss that, and I agree that at the point of my initial lock
break up, the locking is "wrong". Whether you correct it by changing
the lock ordering or by using RCU to do lookups is something I want to
debate further.

I think it is natural to be able to lock the inode and have it lock the
icache state.


> As I've said, I've no religious problems with trylocks; we *do* need them for
> prune_icache() to get a sane locking scheme.  But the way you put ->i_lock on
> the top of hierarchy is simply wrong.

(well that could be avoided with RCU too)

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