From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pf0-f198.google.com (mail-pf0-f198.google.com [209.85.192.198]) by kanga.kvack.org (Postfix) with ESMTP id AEFFE6B0268 for ; Fri, 27 Oct 2017 13:22:07 -0400 (EDT) Received: by mail-pf0-f198.google.com with SMTP id y128so5330620pfg.5 for ; Fri, 27 Oct 2017 10:22:07 -0700 (PDT) Received: from mga05.intel.com (mga05.intel.com. [192.55.52.43]) by mx.google.com with ESMTPS id e13si1287211pgt.166.2017.10.27.10.22.06 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Fri, 27 Oct 2017 10:22:06 -0700 (PDT) Date: Fri, 27 Oct 2017 10:22:05 -0700 From: Andi Kleen Subject: Re: [PATCH] mm: Simplify and batch working set shadow pages LRU isolation locking Message-ID: <20171027172205.GA22894@tassilo.jf.intel.com> References: <20171026234854.25764-1-andi@firstfloor.org> <20171027170156.GA1743@cmpxchg.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20171027170156.GA1743@cmpxchg.org> Sender: owner-linux-mm@kvack.org List-ID: To: Johannes Weiner Cc: Andi Kleen , akpm@linux-foundation.org, linux-mm@kvack.org > The nlru->lock in list_lru_shrink_walk() is the only thing that keeps > truncation blocked on workingset_update_node() -> list_lru_del() and > so ultimately keeping it from freeing the radix tree node. > > It's not safe to access the nodes on the private list after that. True. > > Batching mapping->tree_lock is possible, but you have to keep the > lock-handoff scheme. Pass a &mapping to list_lru_shrink_walk() and > only unlock and spin_trylock(&mapping->tree_lock) if it changes? Yes something like that could work. Thanks. -Andi -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org