From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752286AbdJSJNQ (ORCPT ); Thu, 19 Oct 2017 05:13:16 -0400 Received: from mx2.suse.de ([195.135.220.15]:58887 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751900AbdJSJMz (ORCPT ); Thu, 19 Oct 2017 05:12:55 -0400 Subject: Re: [PATCH 4/8] mm: Only drain per-cpu pagevecs once per pagevec usage To: Mel Gorman , Andrew Morton Cc: Linux-MM , Linux-FSDevel , LKML , Jan Kara , Andi Kleen , Dave Hansen , Dave Chinner References: <20171018075952.10627-1-mgorman@techsingularity.net> <20171018075952.10627-5-mgorman@techsingularity.net> From: Vlastimil Babka Message-ID: Date: Thu, 19 Oct 2017 11:12:52 +0200 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.4.0 MIME-Version: 1.0 In-Reply-To: <20171018075952.10627-5-mgorman@techsingularity.net> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 10/18/2017 09:59 AM, Mel Gorman wrote: > When a pagevec is initialised on the stack, it is generally used multiple > times over a range of pages, looking up entries and then releasing them. > On each pagevec_release, the per-cpu deferred LRU pagevecs are drained > on the grounds the page being released may be on those queues and the > pages may be cache hot. In many cases only the first drain is necessary > as it's unlikely that the range of pages being walked is racing against > LRU addition. Even if there is such a race, the impact is marginal where > as constantly redraining the lru pagevecs costs. Right, the drain is only to a local cpu, not all of them, so that kind of "racing" shouldn't be even possible. > This patch ensures that pagevec is only drained once in a given lifecycle > without increasing the cache footprint of the pagevec structure. Only Well, strictly speaking it does prevent decreasing the cache footprint by removing the 'cold' field later :) > sparsetruncate tiny is shown here as large files have many exceptional > entries and calls pagecache_release less frequently. > > sparsetruncate (tiny) > 4.14.0-rc4 4.14.0-rc4 > batchshadow-v1r1 onedrain-v1r1 > Min Time 141.00 ( 0.00%) 141.00 ( 0.00%) > 1st-qrtle Time 142.00 ( 0.00%) 142.00 ( 0.00%) > 2nd-qrtle Time 142.00 ( 0.00%) 142.00 ( 0.00%) > 3rd-qrtle Time 143.00 ( 0.00%) 143.00 ( 0.00%) > Max-90% Time 144.00 ( 0.00%) 144.00 ( 0.00%) > Max-95% Time 146.00 ( 0.00%) 145.00 ( 0.68%) > Max-99% Time 198.00 ( 0.00%) 194.00 ( 2.02%) > Max Time 254.00 ( 0.00%) 208.00 ( 18.11%) > Amean Time 145.12 ( 0.00%) 144.30 ( 0.56%) > Stddev Time 12.74 ( 0.00%) 9.62 ( 24.49%) > Coeff Time 8.78 ( 0.00%) 6.67 ( 24.06%) > Best99%Amean Time 144.29 ( 0.00%) 143.82 ( 0.32%) > Best95%Amean Time 142.68 ( 0.00%) 142.31 ( 0.26%) > Best90%Amean Time 142.52 ( 0.00%) 142.19 ( 0.24%) > Best75%Amean Time 142.26 ( 0.00%) 141.98 ( 0.20%) > Best50%Amean Time 141.90 ( 0.00%) 141.71 ( 0.13%) > Best25%Amean Time 141.80 ( 0.00%) 141.43 ( 0.26%) > > The impact on bonnie is marginal and within the noise because a significant > percentage of the file being truncated has been reclaimed and consists of > shadow entries which reduce the hotness of the pagevec_release path. > > Signed-off-by: Mel Gorman > --- > include/linux/pagevec.h | 4 +++- > mm/swap.c | 5 ++++- > 2 files changed, 7 insertions(+), 2 deletions(-) > > diff --git a/include/linux/pagevec.h b/include/linux/pagevec.h > index 4dcd5506f1ed..4231979be982 100644 > --- a/include/linux/pagevec.h > +++ b/include/linux/pagevec.h > @@ -16,7 +16,8 @@ struct address_space; > > struct pagevec { > unsigned long nr; > - unsigned long cold; > + bool cold; > + bool drained; 'drained' sounds a bit misleading to me, I would expect it to refer to *this* pagevec. What about e.g. "lru_drained"? Thanks > struct page *pages[PAGEVEC_SIZE]; > }; > > @@ -45,6 +46,7 @@ static inline void pagevec_init(struct pagevec *pvec, int cold) > { > pvec->nr = 0; > pvec->cold = cold; > + pvec->drained = false; > } > > static inline void pagevec_reinit(struct pagevec *pvec) > diff --git a/mm/swap.c b/mm/swap.c > index a77d68f2c1b6..31bd9d8a5db7 100644 > --- a/mm/swap.c > +++ b/mm/swap.c > @@ -833,7 +833,10 @@ EXPORT_SYMBOL(release_pages); > */ > void __pagevec_release(struct pagevec *pvec) > { > - lru_add_drain(); > + if (!pvec->drained) { > + lru_add_drain(); > + pvec->drained = true; > + } > release_pages(pvec->pages, pagevec_count(pvec), pvec->cold); > pagevec_reinit(pvec); > } > From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Subject: Re: [PATCH 4/8] mm: Only drain per-cpu pagevecs once per pagevec usage To: Mel Gorman , Andrew Morton Cc: Linux-MM , Linux-FSDevel , LKML , Jan Kara , Andi Kleen , Dave Hansen , Dave Chinner References: <20171018075952.10627-1-mgorman@techsingularity.net> <20171018075952.10627-5-mgorman@techsingularity.net> From: Vlastimil Babka Message-ID: Date: Thu, 19 Oct 2017 11:12:52 +0200 MIME-Version: 1.0 In-Reply-To: <20171018075952.10627-5-mgorman@techsingularity.net> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: On 10/18/2017 09:59 AM, Mel Gorman wrote: > When a pagevec is initialised on the stack, it is generally used multiple > times over a range of pages, looking up entries and then releasing them. > On each pagevec_release, the per-cpu deferred LRU pagevecs are drained > on the grounds the page being released may be on those queues and the > pages may be cache hot. In many cases only the first drain is necessary > as it's unlikely that the range of pages being walked is racing against > LRU addition. Even if there is such a race, the impact is marginal where > as constantly redraining the lru pagevecs costs. Right, the drain is only to a local cpu, not all of them, so that kind of "racing" shouldn't be even possible. > This patch ensures that pagevec is only drained once in a given lifecycle > without increasing the cache footprint of the pagevec structure. Only Well, strictly speaking it does prevent decreasing the cache footprint by removing the 'cold' field later :) > sparsetruncate tiny is shown here as large files have many exceptional > entries and calls pagecache_release less frequently. > > sparsetruncate (tiny) > 4.14.0-rc4 4.14.0-rc4 > batchshadow-v1r1 onedrain-v1r1 > Min Time 141.00 ( 0.00%) 141.00 ( 0.00%) > 1st-qrtle Time 142.00 ( 0.00%) 142.00 ( 0.00%) > 2nd-qrtle Time 142.00 ( 0.00%) 142.00 ( 0.00%) > 3rd-qrtle Time 143.00 ( 0.00%) 143.00 ( 0.00%) > Max-90% Time 144.00 ( 0.00%) 144.00 ( 0.00%) > Max-95% Time 146.00 ( 0.00%) 145.00 ( 0.68%) > Max-99% Time 198.00 ( 0.00%) 194.00 ( 2.02%) > Max Time 254.00 ( 0.00%) 208.00 ( 18.11%) > Amean Time 145.12 ( 0.00%) 144.30 ( 0.56%) > Stddev Time 12.74 ( 0.00%) 9.62 ( 24.49%) > Coeff Time 8.78 ( 0.00%) 6.67 ( 24.06%) > Best99%Amean Time 144.29 ( 0.00%) 143.82 ( 0.32%) > Best95%Amean Time 142.68 ( 0.00%) 142.31 ( 0.26%) > Best90%Amean Time 142.52 ( 0.00%) 142.19 ( 0.24%) > Best75%Amean Time 142.26 ( 0.00%) 141.98 ( 0.20%) > Best50%Amean Time 141.90 ( 0.00%) 141.71 ( 0.13%) > Best25%Amean Time 141.80 ( 0.00%) 141.43 ( 0.26%) > > The impact on bonnie is marginal and within the noise because a significant > percentage of the file being truncated has been reclaimed and consists of > shadow entries which reduce the hotness of the pagevec_release path. > > Signed-off-by: Mel Gorman > --- > include/linux/pagevec.h | 4 +++- > mm/swap.c | 5 ++++- > 2 files changed, 7 insertions(+), 2 deletions(-) > > diff --git a/include/linux/pagevec.h b/include/linux/pagevec.h > index 4dcd5506f1ed..4231979be982 100644 > --- a/include/linux/pagevec.h > +++ b/include/linux/pagevec.h > @@ -16,7 +16,8 @@ struct address_space; > > struct pagevec { > unsigned long nr; > - unsigned long cold; > + bool cold; > + bool drained; 'drained' sounds a bit misleading to me, I would expect it to refer to *this* pagevec. What about e.g. "lru_drained"? Thanks > struct page *pages[PAGEVEC_SIZE]; > }; > > @@ -45,6 +46,7 @@ static inline void pagevec_init(struct pagevec *pvec, int cold) > { > pvec->nr = 0; > pvec->cold = cold; > + pvec->drained = false; > } > > static inline void pagevec_reinit(struct pagevec *pvec) > diff --git a/mm/swap.c b/mm/swap.c > index a77d68f2c1b6..31bd9d8a5db7 100644 > --- a/mm/swap.c > +++ b/mm/swap.c > @@ -833,7 +833,10 @@ EXPORT_SYMBOL(release_pages); > */ > void __pagevec_release(struct pagevec *pvec) > { > - lru_add_drain(); > + if (!pvec->drained) { > + lru_add_drain(); > + pvec->drained = true; > + } > release_pages(pvec->pages, pagevec_count(pvec), pvec->cold); > pagevec_reinit(pvec); > } > -- 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