Return-Path: linux-nfs-owner@vger.kernel.org Received: from bombadil.infradead.org ([198.137.202.9]:44028 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754116Ab3LDNki (ORCPT ); Wed, 4 Dec 2013 08:40:38 -0500 Date: Wed, 4 Dec 2013 05:40:36 -0800 From: Christoph Hellwig To: Jeff Layton Cc: Christoph Hellwig , bfields@fieldses.org, gartim@gmail.com, linux-nfs@vger.kernel.org Subject: Re: [PATCH] nfsd: when reusing an existing repcache entry, unhash it first Message-ID: <20131204134036.GA3953@infradead.org> References: <1386015979-27511-1-git-send-email-jlayton@redhat.com> <20131203102517.GA12576@infradead.org> <20131203132112.1f19c014@tlielax.poochiereds.net> <20131204083336.GB30216@infradead.org> <20131204075402.7b00d09d@tlielax.poochiereds.net> <20131204130944.GA3464@infradead.org> <20131204083101.6422fb40@tlielax.poochiereds.net> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20131204083101.6422fb40@tlielax.poochiereds.net> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Wed, Dec 04, 2013 at 08:31:01AM -0500, Jeff Layton wrote: > > Exactly. Btw, what about the comment that the DRC should be per-client > > on top of the file? That also seems like some very low hanging fruit. > > > > Ok, I'll see about spinning up a patchset that does that. We'll also > probably want to just allocate entries out of the slabcache in that > case, so that'll make things quite a bit simpler. FYI I'm testing a patch at the moment that just rips out the "direct reclaim" as an alternative to your patch, without actually redoing the whole LRU infrastructure. Unfortunately with your patch 075 still hangs for me, although it at lest doesn't crash.