From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753726AbaHNBOJ (ORCPT ); Wed, 13 Aug 2014 21:14:09 -0400 Received: from lgeamrelo04.lge.com ([156.147.1.127]:62549 "EHLO lgeamrelo04.lge.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752580AbaHNBOI (ORCPT ); Wed, 13 Aug 2014 21:14:08 -0400 X-Original-SENDERIP: 10.177.222.156 X-Original-MAILFROM: minchan.kim@lge.com Date: Thu, 14 Aug 2014 10:14:16 +0900 From: Minchan Kim To: Andrew Morton Cc: Seth Jennings , Jerome Marchand , Sergey Senozhatsky , juno.choi@lge.com, linux-kernel@vger.kernel.org, linux-mm@kvack.org, Dan Streetman , seungho1.park@lge.com, Luigi Semenzato , ds2horner@gmail.com, Nitin Gupta Subject: Re: [PATCH 1/2] zsmalloc: move pages_allocated to zs_pool Message-ID: <20140814011416.GI9227@bbox> References: <1407977877-18185-1-git-send-email-minchan@kernel.org> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <1407977877-18185-1-git-send-email-minchan@kernel.org> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Just resent new one with correcting Seth's mail address and including the one omitted by mistake in this patchset. On Thu, Aug 14, 2014 at 09:57:56AM +0900, Minchan Kim wrote: > Pages_allocated has counted in size_class structure and when user > want to see total_size_bytes, it gathers all of value from each > size_class to report the sum. > > It's not bad if user don't see the value often but if user start > to see the value frequently, it would be not a good deal for > performance POV. > > Even, this patch moves the variable from size_class to zs_pool > so it reduces memory footprint (from [255 * 8byte] to > [sizeof(atomic_t)]) but it introduce new atomic opearation > but it's not a big deal because atomic operation is called on > slow path of zsmalloc where it allocates/free zspage unit, > not object. > > Signed-off-by: Minchan Kim > --- > mm/zsmalloc.c | 21 +++++++-------------- > 1 file changed, 7 insertions(+), 14 deletions(-) > > diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c > index 4e2fc83cb394..2f21ea8921dc 100644 > --- a/mm/zsmalloc.c > +++ b/mm/zsmalloc.c > @@ -199,9 +199,6 @@ struct size_class { > > spinlock_t lock; > > - /* stats */ > - u64 pages_allocated; > - > struct page *fullness_list[_ZS_NR_FULLNESS_GROUPS]; > }; > > @@ -220,6 +217,7 @@ struct zs_pool { > struct size_class size_class[ZS_SIZE_CLASSES]; > > gfp_t flags; /* allocation flags used when growing pool */ > + atomic_t pages_allocated; > }; > > /* > @@ -1027,8 +1025,8 @@ unsigned long zs_malloc(struct zs_pool *pool, size_t size) > return 0; > > set_zspage_mapping(first_page, class->index, ZS_EMPTY); > + atomic_add(class->pages_per_zspage, &pool->pages_allocated); > spin_lock(&class->lock); > - class->pages_allocated += class->pages_per_zspage; > } > > obj = (unsigned long)first_page->freelist; > @@ -1081,14 +1079,12 @@ void zs_free(struct zs_pool *pool, unsigned long obj) > > first_page->inuse--; > fullness = fix_fullness_group(pool, first_page); > - > - if (fullness == ZS_EMPTY) > - class->pages_allocated -= class->pages_per_zspage; > - > spin_unlock(&class->lock); > > - if (fullness == ZS_EMPTY) > + if (fullness == ZS_EMPTY) { > + atomic_sub(class->pages_per_zspage, &pool->pages_allocated); > free_zspage(first_page); > + } > } > EXPORT_SYMBOL_GPL(zs_free); > > @@ -1184,12 +1180,9 @@ EXPORT_SYMBOL_GPL(zs_unmap_object); > > u64 zs_get_total_size_bytes(struct zs_pool *pool) > { > - int i; > - u64 npages = 0; > - > - for (i = 0; i < ZS_SIZE_CLASSES; i++) > - npages += pool->size_class[i].pages_allocated; > + u64 npages; > > + npages = atomic_read(&pool->pages_allocated); > return npages << PAGE_SHIFT; > } > EXPORT_SYMBOL_GPL(zs_get_total_size_bytes); > -- > 2.0.0 -- Kind regards, Minchan Kim