From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by smtp.lore.kernel.org (Postfix) with ESMTP id 49B38C433EF for ; Sat, 19 Feb 2022 16:33:52 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id 54A446B0071; Sat, 19 Feb 2022 11:33:51 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 4FB996B0072; Sat, 19 Feb 2022 11:33:51 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 3C0586B0074; Sat, 19 Feb 2022 11:33:51 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0112.hostedemail.com [216.40.44.112]) by kanga.kvack.org (Postfix) with ESMTP id 292C26B0071 for ; Sat, 19 Feb 2022 11:33:51 -0500 (EST) Received: from smtpin12.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay02.hostedemail.com (Postfix) with ESMTP id D52E29CA14 for ; Sat, 19 Feb 2022 16:33:50 +0000 (UTC) X-FDA: 79160075820.12.ECC8732 Received: from eu-smtp-delivery-151.mimecast.com (eu-smtp-delivery-151.mimecast.com [185.58.86.151]) by imf05.hostedemail.com (Postfix) with ESMTP id 2B263100005 for ; Sat, 19 Feb 2022 16:33:49 +0000 (UTC) Received: from AcuMS.aculab.com (156.67.243.121 [156.67.243.121]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA384) id uk-mta-115-134YVXIbNfWUuHtnEmsqEg-1; Sat, 19 Feb 2022 16:33:46 +0000 X-MC-Unique: 134YVXIbNfWUuHtnEmsqEg-1 Received: from AcuMS.Aculab.com (fd9f:af1c:a25b:0:994c:f5c2:35d6:9b65) by AcuMS.aculab.com (fd9f:af1c:a25b:0:994c:f5c2:35d6:9b65) with Microsoft SMTP Server (TLS) id 15.0.1497.28; Sat, 19 Feb 2022 16:33:45 +0000 Received: from AcuMS.Aculab.com ([fe80::994c:f5c2:35d6:9b65]) by AcuMS.aculab.com ([fe80::994c:f5c2:35d6:9b65%12]) with mapi id 15.00.1497.028; Sat, 19 Feb 2022 16:33:44 +0000 From: David Laight To: 'Miaohe Lin' , "akpm@linux-foundation.org" CC: "vitaly.wool@konsulko.com" , "linux-mm@kvack.org" , "linux-kernel@vger.kernel.org" Subject: RE: [PATCH 6/9] mm/z3fold: move decrement of pool->pages_nr into __release_z3fold_page() Thread-Topic: [PATCH 6/9] mm/z3fold: move decrement of pool->pages_nr into __release_z3fold_page() Thread-Index: AQHYJXLUWbfrUcWr0kay123DGS2luaybEHRg Date: Sat, 19 Feb 2022 16:33:44 +0000 Message-ID: References: <20220219092533.12596-1-linmiaohe@huawei.com> <20220219092533.12596-7-linmiaohe@huawei.com> In-Reply-To: <20220219092533.12596-7-linmiaohe@huawei.com> Accept-Language: en-GB, en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-ms-exchange-transport-fromentityheader: Hosted x-originating-ip: [10.202.205.107] MIME-Version: 1.0 X-Mimecast-Spam-Score: 0 X-Mimecast-Originator: aculab.com Content-Language: en-US Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable X-Rspamd-Queue-Id: 2B263100005 X-Stat-Signature: i6k4rft1s3iuxzryd1u5t8kgmnnndnty Authentication-Results: imf05.hostedemail.com; dkim=none; dmarc=pass (policy=none) header.from=aculab.com; spf=pass (imf05.hostedemail.com: domain of david.laight@aculab.com designates 185.58.86.151 as permitted sender) smtp.mailfrom=david.laight@aculab.com X-Rspam-User: X-Rspamd-Server: rspam11 X-HE-Tag: 1645288429-6123 X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: From: Miaohe Lin > Sent: 19 February 2022 09:26 >=20 > The z3fold will always do atomic64_dec(&pool->pages_nr) when the > __release_z3fold_page() is called. Thus we can move decrement of > pool->pages_nr into __release_z3fold_page() to simplify the code. > Also we can reduce the size of z3fold.o ~1k. > Without this patch: > text=09 data=09 bss=09 dec=09 hex=09filename > 15444=09 1376=09 8=09 16828=09 41bc=09mm/z3fold.o > With this patch: > text=09 data=09 bss=09 dec=09 hex=09filename > 15044=09 1248=09 8=09 16300=09 3fac=09mm/z3fold.o I can't see anything obvious in this patch that would reduce the size much. OTOH there are some large functions that are pointlessly marked 'inline'. Maybe the compiler made a better choice? Although it isn't al all obvious why the 'data' size changes. > Signed-off-by: Miaohe Lin > --- > mm/z3fold.c | 41 ++++++++++++----------------------------- > 1 file changed, 12 insertions(+), 29 deletions(-) >=20 > diff --git a/mm/z3fold.c b/mm/z3fold.c > index adc0b3fa4906..18a697f6fe32 100644 > --- a/mm/z3fold.c > +++ b/mm/z3fold.c > @@ -520,6 +520,8 @@ static void __release_z3fold_page(struct z3fold_heade= r *zhdr, bool locked) > =09list_add(&zhdr->buddy, &pool->stale); > =09queue_work(pool->release_wq, &pool->work); > =09spin_unlock(&pool->stale_lock); > + > +=09atomic64_dec(&pool->pages_nr); Looks like you can move the decrement inside the lock. If you can do the same for the increment you can avoid the expensive locked bus cycle. =09David > } >=20 > static void release_z3fold_page(struct kref *ref) > @@ -737,13 +739,9 @@ static struct z3fold_header *compact_single_buddy(st= ruct z3fold_header *zhdr) > =09return new_zhdr; >=20 > out_fail: > -=09if (new_zhdr) { > -=09=09if (kref_put(&new_zhdr->refcount, release_z3fold_page_locked)) > -=09=09=09atomic64_dec(&pool->pages_nr); > -=09=09else { > -=09=09=09add_to_unbuddied(pool, new_zhdr); > -=09=09=09z3fold_page_unlock(new_zhdr); > -=09=09} > +=09if (new_zhdr && !kref_put(&new_zhdr->refcount, release_z3fold_page_lo= cked)) { > +=09=09add_to_unbuddied(pool, new_zhdr); > +=09=09z3fold_page_unlock(new_zhdr); > =09} > =09return NULL; >=20 > @@ -816,10 +814,8 @@ static void do_compact_page(struct z3fold_header *zh= dr, bool locked) > =09list_del_init(&zhdr->buddy); > =09spin_unlock(&pool->lock); >=20 > -=09if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) { > -=09=09atomic64_dec(&pool->pages_nr); > +=09if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) > =09=09return; > -=09} >=20 > =09if (test_bit(PAGE_STALE, &page->private) || > =09 test_and_set_bit(PAGE_CLAIMED, &page->private)) { > @@ -829,9 +825,7 @@ static void do_compact_page(struct z3fold_header *zhd= r, bool locked) >=20 > =09if (!zhdr->foreign_handles && buddy_single(zhdr) && > =09 zhdr->mapped_count =3D=3D 0 && compact_single_buddy(zhdr)) { > -=09=09if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) > -=09=09=09atomic64_dec(&pool->pages_nr); > -=09=09else { > +=09=09if (!kref_put(&zhdr->refcount, release_z3fold_page_locked)) { > =09=09=09clear_bit(PAGE_CLAIMED, &page->private); > =09=09=09z3fold_page_unlock(zhdr); > =09=09} > @@ -1089,10 +1083,8 @@ static int z3fold_alloc(struct z3fold_pool *pool, = size_t size, gfp_t gfp, > =09=09if (zhdr) { > =09=09=09bud =3D get_free_buddy(zhdr, chunks); > =09=09=09if (bud =3D=3D HEADLESS) { > -=09=09=09=09if (kref_put(&zhdr->refcount, > +=09=09=09=09if (!kref_put(&zhdr->refcount, > =09=09=09=09=09 release_z3fold_page_locked)) > -=09=09=09=09=09atomic64_dec(&pool->pages_nr); > -=09=09=09=09else > =09=09=09=09=09z3fold_page_unlock(zhdr); > =09=09=09=09pr_err("No free chunks in unbuddied\n"); > =09=09=09=09WARN_ON(1); > @@ -1239,10 +1231,8 @@ static void z3fold_free(struct z3fold_pool *pool, = unsigned long handle) >=20 > =09if (!page_claimed) > =09=09free_handle(handle, zhdr); > -=09if (kref_put(&zhdr->refcount, release_z3fold_page_locked_list)) { > -=09=09atomic64_dec(&pool->pages_nr); > +=09if (kref_put(&zhdr->refcount, release_z3fold_page_locked_list)) > =09=09return; > -=09} > =09if (page_claimed) { > =09=09/* the page has not been claimed by us */ > =09=09put_z3fold_header(zhdr); > @@ -1353,9 +1343,7 @@ static int z3fold_reclaim_page(struct z3fold_pool *= pool, unsigned int retries) > =09=09=09=09break; > =09=09=09} > =09=09=09if (!z3fold_page_trylock(zhdr)) { > -=09=09=09=09if (kref_put(&zhdr->refcount, > -=09=09=09=09=09=09release_z3fold_page)) > -=09=09=09=09=09atomic64_dec(&pool->pages_nr); > +=09=09=09=09kref_put(&zhdr->refcount, release_z3fold_page); > =09=09=09=09zhdr =3D NULL; > =09=09=09=09continue; /* can't evict at this point */ > =09=09=09} > @@ -1366,10 +1354,8 @@ static int z3fold_reclaim_page(struct z3fold_pool = *pool, unsigned int retries) > =09=09=09 */ > =09=09=09if (zhdr->foreign_handles || > =09=09=09 test_and_set_bit(PAGE_CLAIMED, &page->private)) { > -=09=09=09=09if (kref_put(&zhdr->refcount, > +=09=09=09=09if (!kref_put(&zhdr->refcount, > =09=09=09=09=09=09release_z3fold_page_locked)) > -=09=09=09=09=09atomic64_dec(&pool->pages_nr); > -=09=09=09=09else > =09=09=09=09=09z3fold_page_unlock(zhdr); > =09=09=09=09zhdr =3D NULL; > =09=09=09=09continue; /* can't evict such page */ > @@ -1447,7 +1433,6 @@ static int z3fold_reclaim_page(struct z3fold_pool *= pool, unsigned int retries) > =09=09=09if (kref_put(&zhdr->refcount, > =09=09=09=09=09release_z3fold_page_locked)) { > =09=09=09=09kmem_cache_free(pool->c_handle, slots); > -=09=09=09=09atomic64_dec(&pool->pages_nr); > =09=09=09=09return 0; > =09=09=09} > =09=09=09/* > @@ -1669,10 +1654,8 @@ static void z3fold_page_putback(struct page *page) > =09if (!list_empty(&zhdr->buddy)) > =09=09list_del_init(&zhdr->buddy); > =09INIT_LIST_HEAD(&page->lru); > -=09if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) { > -=09=09atomic64_dec(&pool->pages_nr); > +=09if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) > =09=09return; > -=09} > =09spin_lock(&pool->lock); > =09list_add(&page->lru, &pool->lru); > =09spin_unlock(&pool->lock); > -- > 2.23.0 - Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1= PT, UK Registration No: 1397386 (Wales)