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 X-Spam-Level: X-Spam-Status: No, score=-16.5 required=3.0 tests=BAYES_00,DKIM_INVALID, DKIM_SIGNED,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 9B01FC63697 for ; Sat, 28 Nov 2020 16:16:19 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 3BA2F2224D for ; Sat, 28 Nov 2020 16:16:19 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=fail reason="signature verification failed" (2048-bit key) header.d=infradead.org header.i=@infradead.org header.b="pm1qv/3p" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 3BA2F2224D Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=lst.de Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 8C0BC8D001C; Sat, 28 Nov 2020 11:15:58 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 81D4B8D0017; Sat, 28 Nov 2020 11:15:58 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 5FEF48D001C; Sat, 28 Nov 2020 11:15:58 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0107.hostedemail.com [216.40.44.107]) by kanga.kvack.org (Postfix) with ESMTP id 3FF808D0017 for ; Sat, 28 Nov 2020 11:15:58 -0500 (EST) Received: from smtpin01.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay03.hostedemail.com (Postfix) with ESMTP id 01AA28249980 for ; Sat, 28 Nov 2020 16:15:58 +0000 (UTC) X-FDA: 77534328396.01.order72_2c17be627392 Received: from filter.hostedemail.com (10.5.16.251.rfc1918.com [10.5.16.251]) by smtpin01.hostedemail.com (Postfix) with ESMTP id D840A10051D1C for ; Sat, 28 Nov 2020 16:15:57 +0000 (UTC) X-HE-Tag: order72_2c17be627392 X-Filterd-Recvd-Size: 10624 Received: from casper.infradead.org (casper.infradead.org [90.155.50.34]) by imf11.hostedemail.com (Postfix) with ESMTP for ; Sat, 28 Nov 2020 16:15:57 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=infradead.org; s=casper.20170209; h=Content-Transfer-Encoding:MIME-Version: References:In-Reply-To:Message-Id:Date:Subject:Cc:To:From:Sender:Reply-To: Content-Type:Content-ID:Content-Description; bh=vHZKilZd0flA+9QmXJqftcP+lHbmEPxCO75tRZ/zDDk=; b=pm1qv/3prdcgaI0NAiS7TVFLDy h4ZjV5Alw4Znz3wUC9Xcc64kHzDCE6RBYJS6xAJR+M/FpkdveCqN+ewUrDY0idpjv1UCxnmx82WLn w5kMwGSM8X6afZs8Ftqsy3T8VZ2uNrmdCFT1nBZoBjjXINCzfl0IYk9z9fd0HBzqcm5kXB1bAz6rz T74/H7ssrDgC2+P/8RTyT1zeehuVHYsWm78gIO690EwDeCRe+MB0oR0tY5H9IHv9sUcIaurQjxxpo vftvcXdcMRV1WX+pN5LzsmHFvuBRXernx1sCyinIi/zkHX6FOUsuyyI3R8sUoZ/D2BPFFTlu0YdME DIZ0mLdA==; Received: from [2001:4bb8:18c:1dd6:48f3:741a:602e:7fdd] (helo=localhost) by casper.infradead.org with esmtpsa (Exim 4.92.3 #3 (Red Hat Linux)) id 1kj2t0-0000IS-Is; Sat, 28 Nov 2020 16:15:47 +0000 From: Christoph Hellwig To: Jens Axboe Cc: Tejun Heo , Josef Bacik , Coly Li , Mike Snitzer , Greg Kroah-Hartman , Jan Kara , Johannes Thumshirn , dm-devel@redhat.com, Jan Kara , linux-block@vger.kernel.org, linux-bcache@vger.kernel.org, linux-mtd@lists.infradead.org, linux-fsdevel@vger.kernel.org, linux-mm@kvack.org Subject: [PATCH 21/45] block: refactor blkdev_get Date: Sat, 28 Nov 2020 17:14:46 +0100 Message-Id: <20201128161510.347752-22-hch@lst.de> X-Mailer: git-send-email 2.29.2 In-Reply-To: <20201128161510.347752-1-hch@lst.de> References: <20201128161510.347752-1-hch@lst.de> MIME-Version: 1.0 X-SRS-Rewrite: SMTP reverse-path rewritten from by casper.infradead.org. See http://www.infradead.org/rpr.html Content-Transfer-Encoding: quoted-printable 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: Move more code that is only run on the outer open but not the open of the underlying whole device when opening a partition into blkdev_get, which leads to a much easier to follow structure. This allows to simplify the disk and module refcounting so that one reference is held for each open, similar to what we do with normal file operations. Signed-off-by: Christoph Hellwig Acked-by: Tejun Heo Reviewed-by: Jan Kara --- fs/block_dev.c | 185 +++++++++++++++++++++++-------------------------- 1 file changed, 86 insertions(+), 99 deletions(-) diff --git a/fs/block_dev.c b/fs/block_dev.c index 6016777b648336..0c533ac92e2492 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -1407,46 +1407,12 @@ EXPORT_SYMBOL_GPL(bdev_disk_changed); * mutex_lock(part->bd_mutex) * mutex_lock_nested(whole->bd_mutex, 1) */ - -static int __blkdev_get(struct block_device *bdev, fmode_t mode, void *h= older, - int for_part) +static int __blkdev_get(struct block_device *bdev, struct gendisk *disk, + int partno, fmode_t mode) { - struct block_device *whole =3D NULL, *claiming =3D NULL; - struct gendisk *disk; int ret; - int partno; - bool first_open =3D false, unblock_events =3D true, need_restart; - - restart: - need_restart =3D false; - ret =3D -ENXIO; - disk =3D bdev_get_gendisk(bdev, &partno); - if (!disk) - goto out; - - if (partno) { - whole =3D bdget_disk(disk, 0); - if (!whole) { - ret =3D -ENOMEM; - goto out_put_disk; - } - } =20 - if (!for_part && (mode & FMODE_EXCL)) { - WARN_ON_ONCE(!holder); - if (whole) - claiming =3D whole; - else - claiming =3D bdev; - ret =3D bd_prepare_to_claim(bdev, claiming, holder); - if (ret) - goto out_put_whole; - } - - disk_block_events(disk); - mutex_lock_nested(&bdev->bd_mutex, for_part); if (!bdev->bd_openers) { - first_open =3D true; bdev->bd_disk =3D disk; bdev->bd_contains =3D bdev; bdev->bd_partno =3D partno; @@ -1458,15 +1424,8 @@ static int __blkdev_get(struct block_device *bdev,= fmode_t mode, void *holder, goto out_clear; =20 ret =3D 0; - if (disk->fops->open) { + if (disk->fops->open) ret =3D disk->fops->open(bdev, mode); - /* - * If we lost a race with 'disk' being deleted, - * try again. See md.c - */ - if (ret =3D=3D -ERESTARTSYS) - need_restart =3D true; - } =20 if (!ret) { bd_set_nr_sectors(bdev, get_capacity(disk)); @@ -1486,14 +1445,23 @@ static int __blkdev_get(struct block_device *bdev= , fmode_t mode, void *holder, if (ret) goto out_clear; } else { - BUG_ON(for_part); - ret =3D __blkdev_get(whole, mode, NULL, 1); - if (ret) + struct block_device *whole =3D bdget_disk(disk, 0); + + mutex_lock_nested(&whole->bd_mutex, 1); + ret =3D __blkdev_get(whole, disk, 0, mode); + if (ret) { + mutex_unlock(&whole->bd_mutex); + bdput(whole); goto out_clear; - bdev->bd_contains =3D bdgrab(whole); + } + whole->bd_part_count++; + mutex_unlock(&whole->bd_mutex); + + bdev->bd_contains =3D whole; bdev->bd_part =3D disk_get_part(disk, partno); if (!(disk->flags & GENHD_FL_UP) || !bdev->bd_part || !bdev->bd_part->nr_sects) { + __blkdev_put(whole, mode, 1); ret =3D -ENXIO; goto out_clear; } @@ -1513,58 +1481,17 @@ static int __blkdev_get(struct block_device *bdev= , fmode_t mode, void *holder, (!ret || ret =3D=3D -ENOMEDIUM)) bdev_disk_changed(bdev, ret =3D=3D -ENOMEDIUM); if (ret) - goto out_unlock_bdev; + return ret; } } bdev->bd_openers++; - if (for_part) - bdev->bd_part_count++; - if (claiming) - bd_finish_claiming(bdev, claiming, holder); - - /* - * Block event polling for write claims if requested. Any write holder - * makes the write_holder state stick until all are released. This is - * good enough and tracking individual writeable reference is too - * fragile given the way @mode is used in blkdev_get/put(). - */ - if (claiming && (mode & FMODE_WRITE) && !bdev->bd_write_holder && - (disk->flags & GENHD_FL_BLOCK_EVENTS_ON_EXCL_WRITE)) { - bdev->bd_write_holder =3D true; - unblock_events =3D false; - } - mutex_unlock(&bdev->bd_mutex); - - if (unblock_events) - disk_unblock_events(disk); - - /* only one opener holds refs to the module and disk */ - if (!first_open) - put_disk_and_module(disk); - if (whole) - bdput(whole); return 0; =20 out_clear: disk_put_part(bdev->bd_part); bdev->bd_disk =3D NULL; bdev->bd_part =3D NULL; - if (bdev !=3D bdev->bd_contains) - __blkdev_put(bdev->bd_contains, mode, 1); bdev->bd_contains =3D NULL; - out_unlock_bdev: - if (claiming) - bd_abort_claiming(bdev, claiming, holder); - mutex_unlock(&bdev->bd_mutex); - disk_unblock_events(disk); - out_put_whole: - if (whole) - bdput(whole); - out_put_disk: - put_disk_and_module(disk); - if (need_restart) - goto restart; - out: return ret; } =20 @@ -1589,7 +1516,12 @@ static int __blkdev_get(struct block_device *bdev,= fmode_t mode, void *holder, */ static int blkdev_get(struct block_device *bdev, fmode_t mode, void *hol= der) { - int ret, perm =3D 0; + struct block_device *claiming; + bool unblock_events =3D true; + struct gendisk *disk; + int perm =3D 0; + int partno; + int ret; =20 if (mode & FMODE_READ) perm |=3D MAY_READ; @@ -1599,13 +1531,67 @@ static int blkdev_get(struct block_device *bdev, = fmode_t mode, void *holder) if (ret) goto bdput; =20 - ret =3D__blkdev_get(bdev, mode, holder, 0); - if (ret) + /* + * If we lost a race with 'disk' being deleted, try again. See md.c. + */ +retry: + ret =3D -ENXIO; + disk =3D bdev_get_gendisk(bdev, &partno); + if (!disk) goto bdput; - return 0; =20 + if (mode & FMODE_EXCL) { + WARN_ON_ONCE(!holder); +=09 + ret =3D -ENOMEM; + claiming =3D bdget_disk(disk, 0); + if (!claiming) + goto put_disk; + ret =3D bd_prepare_to_claim(bdev, claiming, holder); + if (ret) + goto put_claiming; + } + + disk_block_events(disk); + + mutex_lock(&bdev->bd_mutex); + ret =3D__blkdev_get(bdev, disk, partno, mode); + if (!(mode & FMODE_EXCL)) { + ; /* nothing to do here */ + } else if (ret) { + bd_abort_claiming(bdev, claiming, holder); + } else { + bd_finish_claiming(bdev, claiming, holder); + + /* + * Block event polling for write claims if requested. Any write + * holder makes the write_holder state stick until all are + * released. This is good enough and tracking individual + * writeable reference is too fragile given the way @mode is + * used in blkdev_get/put(). + */ + if ((mode & FMODE_WRITE) && !bdev->bd_write_holder && + (disk->flags & GENHD_FL_BLOCK_EVENTS_ON_EXCL_WRITE)) { + bdev->bd_write_holder =3D true; + unblock_events =3D false; + } + } + mutex_unlock(&bdev->bd_mutex); + + if (unblock_events) + disk_unblock_events(disk); + +put_claiming: + if (mode & FMODE_EXCL) + bdput(claiming); +put_disk: + if (ret) + put_disk_and_module(disk); + if (ret =3D=3D -ERESTARTSYS) + goto retry; bdput: - bdput(bdev); + if (ret) + bdput(bdev); return ret; } =20 @@ -1753,8 +1739,6 @@ static void __blkdev_put(struct block_device *bdev,= fmode_t mode, int for_part) if (bdev_is_partition(bdev)) victim =3D bdev->bd_contains; bdev->bd_contains =3D NULL; - - put_disk_and_module(disk); } else { if (!bdev_is_partition(bdev) && disk->fops->release) disk->fops->release(disk, mode); @@ -1767,6 +1751,8 @@ static void __blkdev_put(struct block_device *bdev,= fmode_t mode, int for_part) =20 void blkdev_put(struct block_device *bdev, fmode_t mode) { + struct gendisk *disk =3D bdev->bd_disk; + mutex_lock(&bdev->bd_mutex); =20 if (mode & FMODE_EXCL) { @@ -1795,7 +1781,7 @@ void blkdev_put(struct block_device *bdev, fmode_t = mode) * unblock evpoll if it was a write holder. */ if (bdev_free && bdev->bd_write_holder) { - disk_unblock_events(bdev->bd_disk); + disk_unblock_events(disk); bdev->bd_write_holder =3D false; } } @@ -1805,11 +1791,12 @@ void blkdev_put(struct block_device *bdev, fmode_= t mode) * event. This is to ensure detection of media removal commanded * from userland - e.g. eject(1). */ - disk_flush_events(bdev->bd_disk, DISK_EVENT_MEDIA_CHANGE); + disk_flush_events(disk, DISK_EVENT_MEDIA_CHANGE); =20 mutex_unlock(&bdev->bd_mutex); =20 __blkdev_put(bdev, mode, 0); + put_disk_and_module(disk); } EXPORT_SYMBOL(blkdev_put); =20 --=20 2.29.2