From patchwork Wed Feb 22 21:58:26 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jonathan Derrick X-Patchwork-Id: 13149620 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 83A5CC636D6 for ; Wed, 22 Feb 2023 22:01:47 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230297AbjBVWBq (ORCPT ); Wed, 22 Feb 2023 17:01:46 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:55916 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229604AbjBVWBo (ORCPT ); Wed, 22 Feb 2023 17:01:44 -0500 Received: from resqmta-a1p-077437.sys.comcast.net (resqmta-a1p-077437.sys.comcast.net [IPv6:2001:558:fd01:2bb4::8]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id DD3B241B6D for ; Wed, 22 Feb 2023 14:01:42 -0800 (PST) Received: from resomta-a1p-077058.sys.comcast.net ([96.103.145.239]) by resqmta-a1p-077437.sys.comcast.net with ESMTP id Up31pfZpOKJmTUx8opv8t6; Wed, 22 Feb 2023 21:59:10 +0000 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=comcastmailservice.net; s=20211018a; t=1677103150; bh=XvX+8ukf+hGDaGzInSZ4GlkxIDagerSJ2nydRL7t5Ks=; h=Received:Received:From:To:Subject:Date:Message-Id:MIME-Version: Xfinity-Spam-Result; b=ouvINojGpofH8zV+izjl3aADHxcg4vDWJFa170GMNdVAVo8iUarXfobFtKk9hIF4F LEuO/qEY9p2wl6YyRqp/j67DNWgaLOi9cnM+dsw/vcbovrghvFazxqsmJ2RBL6GQ9g s2+YavhmjsglSjELANkxELlfShBsWVGNvf7JMP4XSEnGp81rOuC/IP6sOpzlqgc4EP Of43w/gZXCLUCvdTOIMQBD+xqzntHOOyMWtyOvy1lK5swZIVB5kIQFNiF/fRYm2gXQ s5HIcyZKktD+t67dJtOTutjrzq69R8kUx4Y9P78S17IafATqM1Ru7/o4aqhetIR8It X0rNgLUBKF85w== Received: from jderrick-mobl4.amr.corp.intel.com ([71.205.181.50]) by resomta-a1p-077058.sys.comcast.net with ESMTPA id Ux8JpSr0ZqdkEUx8Tpx9bM; Wed, 22 Feb 2023 21:58:49 +0000 X-Xfinity-VAAS: gggruggvucftvghtrhhoucdtuddrgedvhedrudejledgudehfecutefuodetggdotefrodftvfcurfhrohhfihhlvgemucevohhmtggrshhtqdftvghsihdpqfgfvfdppffquffrtefokffrnecuuegrihhlohhuthemuceftddunecusecvtfgvtghiphhivghnthhsucdlqddutddtmdenucfjughrpefhvfevufffkffojghfggfgsedtkeertdertddtnecuhfhrohhmpeflohhnrghthhgrnhcuffgvrhhrihgtkhcuoehjohhnrghthhgrnhdruggvrhhrihgtkheslhhinhhugidruggvvheqnecuggftrfgrthhtvghrnheptdetleejfffgffevhefhteevfeeuvdehveffffehtdejuedvvefgfedttdehfedtnecukfhppeejuddrvddthedrudekuddrhedtnecuvehluhhsthgvrhfuihiivgeptdenucfrrghrrghmpehhvghlohepjhguvghrrhhitghkqdhmohgslhegrdgrmhhrrdgtohhrphdrihhnthgvlhdrtghomhdpihhnvghtpeejuddrvddthedrudekuddrhedtpdhmrghilhhfrhhomhepjhhonhgrthhhrghnrdguvghrrhhitghksehlihhnuhigrdguvghvpdhnsggprhgtphhtthhopeejpdhrtghpthhtohepshhonhhgsehkvghrnhgvlhdrohhrghdprhgtphhtthhopehlihhnuhigqdhrrghiugesvhhgvghrrdhkvghrnhgvlhdrohhrghdprhgtphhtthhopeignhhisehrvgguhhgrthdrtghomhdprhgtphhtthhopehhtghhsehinhhfrhgruggvrggurdhorhhgpdhrtghpthhtohepphhmvghnii gvlhesmhholhhgvghnrdhmphhgrdguvgdprhgtphhtthhopehsuhhshhhmrgdrkhgrlhgrkhhothgrsehinhhtvghlrdgtohhm X-Xfinity-VMeta: sc=-100.00;st=legit From: Jonathan Derrick To: Song Liu , Cc: Xiao Ni , Christoph Hellwig , Paul Menzel , Sushma Kalakota , Jon Derrick Subject: [PATCH v3 1/3] md: Move sb writer loop to its own function Date: Wed, 22 Feb 2023 14:58:26 -0700 Message-Id: <20230222215828.225-2-jonathan.derrick@linux.dev> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20230222215828.225-1-jonathan.derrick@linux.dev> References: <20230222215828.225-1-jonathan.derrick@linux.dev> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-raid@vger.kernel.org From: Jon Derrick Preparatory patch for optimal I/O size calculation. Move the sb writer loop routine into its own function for clarity. Signed-off-by: Jon Derrick Reviewed-by: Christoph Hellwig --- drivers/md/md-bitmap.c | 123 +++++++++++++++++++++-------------------- 1 file changed, 63 insertions(+), 60 deletions(-) diff --git a/drivers/md/md-bitmap.c b/drivers/md/md-bitmap.c index e7cc6ba1b657..5c65268a2d97 100644 --- a/drivers/md/md-bitmap.c +++ b/drivers/md/md-bitmap.c @@ -209,76 +209,79 @@ static struct md_rdev *next_active_rdev(struct md_rdev *rdev, struct mddev *mdde return NULL; } -static int write_sb_page(struct bitmap *bitmap, struct page *page, int wait) +static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap, + struct page *page) { - struct md_rdev *rdev; struct block_device *bdev; struct mddev *mddev = bitmap->mddev; struct bitmap_storage *store = &bitmap->storage; + loff_t offset = mddev->bitmap_info.offset; + int size = PAGE_SIZE; + + bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev; + if (page->index == store->file_pages - 1) { + int last_page_size = store->bytes & (PAGE_SIZE - 1); + if (last_page_size == 0) + last_page_size = PAGE_SIZE; + size = roundup(last_page_size, + bdev_logical_block_size(bdev)); + } + + /* Just make sure we aren't corrupting data or metadata */ + if (mddev->external) { + /* Bitmap could be anywhere. */ + if (rdev->sb_start + offset + + (page->index * (PAGE_SIZE / SECTOR_SIZE)) + > rdev->data_offset && + rdev->sb_start + offset + < (rdev->data_offset + mddev->dev_sectors + + (PAGE_SIZE / SECTOR_SIZE))) + return -EINVAL; + } else if (offset < 0) { + /* DATA BITMAP METADATA */ + if (offset + + (long)(page->index * (PAGE_SIZE / SECTOR_SIZE)) + + size / SECTOR_SIZE > 0) + /* bitmap runs in to metadata */ + return -EINVAL; + + if (rdev->data_offset + mddev->dev_sectors + > rdev->sb_start + offset) + /* data runs in to bitmap */ + return -EINVAL; + } else if (rdev->sb_start < rdev->data_offset) { + /* METADATA BITMAP DATA */ + if (rdev->sb_start + offset + + page->index * (PAGE_SIZE / SECTOR_SIZE) + + size / SECTOR_SIZE > rdev->data_offset) + /* bitmap runs in to data */ + return -EINVAL; + } else { + /* DATA METADATA BITMAP - no problems */ + } -restart: - rdev = NULL; - while ((rdev = next_active_rdev(rdev, mddev)) != NULL) { - int size = PAGE_SIZE; - loff_t offset = mddev->bitmap_info.offset; + md_super_write(mddev, rdev, + rdev->sb_start + offset + + page->index * (PAGE_SIZE / SECTOR_SIZE), + size, page); + return 0; +} - bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev; +static int write_sb_page(struct bitmap *bitmap, struct page *page, int wait) +{ + struct md_rdev *rdev; + struct mddev *mddev = bitmap->mddev; + int ret; - if (page->index == store->file_pages-1) { - int last_page_size = store->bytes & (PAGE_SIZE-1); - if (last_page_size == 0) - last_page_size = PAGE_SIZE; - size = roundup(last_page_size, - bdev_logical_block_size(bdev)); - } - /* Just make sure we aren't corrupting data or - * metadata - */ - if (mddev->external) { - /* Bitmap could be anywhere. */ - if (rdev->sb_start + offset + (page->index - * (PAGE_SIZE/512)) - > rdev->data_offset - && - rdev->sb_start + offset - < (rdev->data_offset + mddev->dev_sectors - + (PAGE_SIZE/512))) - goto bad_alignment; - } else if (offset < 0) { - /* DATA BITMAP METADATA */ - if (offset - + (long)(page->index * (PAGE_SIZE/512)) - + size/512 > 0) - /* bitmap runs in to metadata */ - goto bad_alignment; - if (rdev->data_offset + mddev->dev_sectors - > rdev->sb_start + offset) - /* data runs in to bitmap */ - goto bad_alignment; - } else if (rdev->sb_start < rdev->data_offset) { - /* METADATA BITMAP DATA */ - if (rdev->sb_start - + offset - + page->index*(PAGE_SIZE/512) + size/512 - > rdev->data_offset) - /* bitmap runs in to data */ - goto bad_alignment; - } else { - /* DATA METADATA BITMAP - no problems */ + do { + while ((rdev = next_active_rdev(rdev, mddev)) != NULL) { + ret = __write_sb_page(rdev, bitmap, page); + if (ret) + return ret; } - md_super_write(mddev, rdev, - rdev->sb_start + offset - + page->index * (PAGE_SIZE/512), - size, - page); - } + } while (wait && md_super_wait(mddev) < 0); - if (wait && md_super_wait(mddev) < 0) - goto restart; return 0; - - bad_alignment: - return -EINVAL; } static void md_bitmap_file_kick(struct bitmap *bitmap); From patchwork Wed Feb 22 21:58:27 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jonathan Derrick X-Patchwork-Id: 13149619 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 3A082C61DA4 for ; Wed, 22 Feb 2023 22:01:46 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230073AbjBVWBp (ORCPT ); Wed, 22 Feb 2023 17:01:45 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:55914 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229600AbjBVWBo (ORCPT ); Wed, 22 Feb 2023 17:01:44 -0500 X-Greylist: delayed 150 seconds by postgrey-1.37 at lindbergh.monkeyblade.net; Wed, 22 Feb 2023 14:01:42 PST Received: from resqmta-a1p-077436.sys.comcast.net (resqmta-a1p-077436.sys.comcast.net [IPv6:2001:558:fd01:2bb4::2]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id D433841B59 for ; Wed, 22 Feb 2023 14:01:42 -0800 (PST) Received: from resomta-a1p-077058.sys.comcast.net ([96.103.145.239]) by resqmta-a1p-077436.sys.comcast.net with ESMTP id UnbDpYZ1gi3AlUx8opEVPo; Wed, 22 Feb 2023 21:59:10 +0000 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=comcastmailservice.net; s=20211018a; t=1677103150; bh=qPdLhx5ycWaT4aMCe5WDiXmSIfDtmuyTn4IKLqHHlWY=; h=Received:Received:From:To:Subject:Date:Message-Id:MIME-Version: Xfinity-Spam-Result; b=vV9KnGp3GAYjtb4Hy0pvZ2chgpgHULnE8h+ss61BHdyxxHULXddfUne1I2B1CH5Xe UdSvATLHtwwd76jVn39GX4wJiqJWd5KiV73xOq5mCwyIlWzxM0OeoW9fXKS6hI43Rl n1f2q0uX/r2lDYhxFLXHswaHValM/MRxMnsKYm81FN/zcNPL2ysWi7btICup3R4OaX GNYlc7xWnSPnU9lWakJnMIk9qOdPHk8A0XO/qDfNgMXIuyB4NUYxicnTGDhWkRyWEi KMiZgHUcL/o945+NfI3IU5HYH4K8MYUbGKeyJ82WPlHOiHctSl2TUwLgHdEdJED1GD 1G276IR66qo/g== Received: from jderrick-mobl4.amr.corp.intel.com ([71.205.181.50]) by resomta-a1p-077058.sys.comcast.net with ESMTPA id Ux8JpSr0ZqdkEUx8Upx9bW; Wed, 22 Feb 2023 21:58:50 +0000 X-Xfinity-VAAS: gggruggvucftvghtrhhoucdtuddrgedvhedrudejledgudehfecutefuodetggdotefrodftvfcurfhrohhfihhlvgemucevohhmtggrshhtqdftvghsihdpqfgfvfdppffquffrtefokffrnecuuegrihhlohhuthemuceftddunecusecvtfgvtghiphhivghnthhsucdlqddutddtmdenucfjughrpefhvfevufffkffojghfggfgsedtkeertdertddtnecuhfhrohhmpeflohhnrghthhgrnhcuffgvrhhrihgtkhcuoehjohhnrghthhgrnhdruggvrhhrihgtkheslhhinhhugidruggvvheqnecuggftrfgrthhtvghrnheptdetleejfffgffevhefhteevfeeuvdehveffffehtdejuedvvefgfedttdehfedtnecukfhppeejuddrvddthedrudekuddrhedtnecuvehluhhsthgvrhfuihiivgeptdenucfrrghrrghmpehhvghlohepjhguvghrrhhitghkqdhmohgslhegrdgrmhhrrdgtohhrphdrihhnthgvlhdrtghomhdpihhnvghtpeejuddrvddthedrudekuddrhedtpdhmrghilhhfrhhomhepjhhonhgrthhhrghnrdguvghrrhhitghksehlihhnuhigrdguvghvpdhnsggprhgtphhtthhopeejpdhrtghpthhtohepshhonhhgsehkvghrnhgvlhdrohhrghdprhgtphhtthhopehlihhnuhigqdhrrghiugesvhhgvghrrdhkvghrnhgvlhdrohhrghdprhgtphhtthhopeignhhisehrvgguhhgrthdrtghomhdprhgtphhtthhopehhtghhsehinhhfrhgruggvrggurdhorhhgpdhrtghpthhtohepphhmvghnii gvlhesmhholhhgvghnrdhmphhgrdguvgdprhgtphhtthhopehsuhhshhhmrgdrkhgrlhgrkhhothgrsehinhhtvghlrdgtohhm X-Xfinity-VMeta: sc=-100.00;st=legit From: Jonathan Derrick To: Song Liu , Cc: Xiao Ni , Christoph Hellwig , Paul Menzel , Sushma Kalakota , Jon Derrick Subject: [PATCH v3 2/3] md: Fix types in sb writer Date: Wed, 22 Feb 2023 14:58:27 -0700 Message-Id: <20230222215828.225-3-jonathan.derrick@linux.dev> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20230222215828.225-1-jonathan.derrick@linux.dev> References: <20230222215828.225-1-jonathan.derrick@linux.dev> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-raid@vger.kernel.org From: Jon Derrick Page->index is a pgoff_t and multiplying could cause overflows on a 32-bit architecture. In the sb writer, this is used to calculate and verify the sector being used, and is multiplied by a sector value. Using sector_t will cast it to a u64 type and is the more appropriate type for the unit. Additionally, the integer size unit is converted to a sector unit in later calculations, and is now corrected to be an unsigned type. Finally, clean up the calculations using variable aliases to improve readabiliy. Signed-off-by: Jon Derrick Reviewed-by: Christoph Hellwig --- drivers/md/md-bitmap.c | 36 +++++++++++++++--------------------- 1 file changed, 15 insertions(+), 21 deletions(-) diff --git a/drivers/md/md-bitmap.c b/drivers/md/md-bitmap.c index 5c65268a2d97..11f4453775ee 100644 --- a/drivers/md/md-bitmap.c +++ b/drivers/md/md-bitmap.c @@ -215,55 +215,49 @@ static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap, struct block_device *bdev; struct mddev *mddev = bitmap->mddev; struct bitmap_storage *store = &bitmap->storage; - loff_t offset = mddev->bitmap_info.offset; - int size = PAGE_SIZE; + sector_t offset = mddev->bitmap_info.offset; + sector_t ps, sboff, doff; + unsigned int size = PAGE_SIZE; bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev; if (page->index == store->file_pages - 1) { - int last_page_size = store->bytes & (PAGE_SIZE - 1); + unsigned int last_page_size = store->bytes & (PAGE_SIZE - 1); if (last_page_size == 0) last_page_size = PAGE_SIZE; size = roundup(last_page_size, bdev_logical_block_size(bdev)); } + ps = page->index * (PAGE_SIZE / SECTOR_SIZE); + sboff = rdev->sb_start + offset; + doff = rdev->data_offset; + /* Just make sure we aren't corrupting data or metadata */ if (mddev->external) { /* Bitmap could be anywhere. */ - if (rdev->sb_start + offset - + (page->index * (PAGE_SIZE / SECTOR_SIZE)) - > rdev->data_offset && - rdev->sb_start + offset - < (rdev->data_offset + mddev->dev_sectors - + (PAGE_SIZE / SECTOR_SIZE))) + if (sboff + ps > doff && + sboff < (doff + mddev->dev_sectors + (PAGE_SIZE / SECTOR_SIZE))) return -EINVAL; } else if (offset < 0) { /* DATA BITMAP METADATA */ - if (offset - + (long)(page->index * (PAGE_SIZE / SECTOR_SIZE)) - + size / SECTOR_SIZE > 0) + if (offset + ps + size / SECTOR_SIZE > 0) /* bitmap runs in to metadata */ return -EINVAL; - if (rdev->data_offset + mddev->dev_sectors - > rdev->sb_start + offset) + if (doff + mddev->dev_sectors > sboff) /* data runs in to bitmap */ return -EINVAL; } else if (rdev->sb_start < rdev->data_offset) { /* METADATA BITMAP DATA */ - if (rdev->sb_start + offset - + page->index * (PAGE_SIZE / SECTOR_SIZE) - + size / SECTOR_SIZE > rdev->data_offset) + if (sboff + ps + size / SECTOR_SIZE > doff) /* bitmap runs in to data */ return -EINVAL; } else { /* DATA METADATA BITMAP - no problems */ } - md_super_write(mddev, rdev, - rdev->sb_start + offset - + page->index * (PAGE_SIZE / SECTOR_SIZE), - size, page); + md_super_write(mddev, rdev, sboff + ps, + (int) size, page); return 0; } From patchwork Wed Feb 22 21:58:28 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jonathan Derrick X-Patchwork-Id: 13149622 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 80B4FC636D6 for ; Wed, 22 Feb 2023 22:01:50 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229755AbjBVWBs (ORCPT ); Wed, 22 Feb 2023 17:01:48 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:55920 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230114AbjBVWBp (ORCPT ); Wed, 22 Feb 2023 17:01:45 -0500 Received: from resqmta-a1p-077724.sys.comcast.net (resqmta-a1p-077724.sys.comcast.net [IPv6:2001:558:fd01:2bb4::5]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id DC1DC41B5F for ; Wed, 22 Feb 2023 14:01:42 -0800 (PST) Received: from resomta-a1p-077058.sys.comcast.net ([96.103.145.239]) by resqmta-a1p-077724.sys.comcast.net with ESMTP id Uphjpj5j4S9f3Ux8op6aZn; Wed, 22 Feb 2023 21:59:10 +0000 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=comcastmailservice.net; s=20211018a; t=1677103150; bh=XTRLMoVLaadCgZz6CHw3Me+EomanXlgSJsgcV3FeuiU=; h=Received:Received:From:To:Subject:Date:Message-Id:MIME-Version: Xfinity-Spam-Result; b=r+SguYkxGKhPZUJKQrl0q6J8cdNl0aPP9kUvccPsGBvB3xbjevdK9IJ9Br5ugH59d TXTmeJlefOMbD0as8Gm6HuMHcqNnMxKosv0AX8gd7fk/SvZheYgbAi8+FGdUl+Tp5F LDOI8JKqhlXqi7ldI0dLtHjDRekngQkBXS1vMjEvC91mCgXhbmEeqUi90k6kS9wpEm LhdPf36sO2ILScZiol3M/NmG3T7ATNEkMK8xfv3Yn5256gIWukjpAox82//96byi/X eIshoOE/Vn0L5LZtIjQj8CUiVAZSAt/FOvpA9CDTpbH9eppOz0oygId/qLaO5+tXWS Wi1caYRxsm0PA== Received: from jderrick-mobl4.amr.corp.intel.com ([71.205.181.50]) by resomta-a1p-077058.sys.comcast.net with ESMTPA id Ux8JpSr0ZqdkEUx8Vpx9bc; Wed, 22 Feb 2023 21:58:51 +0000 X-Xfinity-VAAS: gggruggvucftvghtrhhoucdtuddrgedvhedrudejledgudehfecutefuodetggdotefrodftvfcurfhrohhfihhlvgemucevohhmtggrshhtqdftvghsihdpqfgfvfdppffquffrtefokffrnecuuegrihhlohhuthemuceftddunecusecvtfgvtghiphhivghnthhsucdlqddutddtmdenucfjughrpefhvfevufffkffojghfggfgsedtkeertdertddtnecuhfhrohhmpeflohhnrghthhgrnhcuffgvrhhrihgtkhcuoehjohhnrghthhgrnhdruggvrhhrihgtkheslhhinhhugidruggvvheqnecuggftrfgrthhtvghrnheptdetleejfffgffevhefhteevfeeuvdehveffffehtdejuedvvefgfedttdehfedtnecukfhppeejuddrvddthedrudekuddrhedtnecuvehluhhsthgvrhfuihiivgeptdenucfrrghrrghmpehhvghlohepjhguvghrrhhitghkqdhmohgslhegrdgrmhhrrdgtohhrphdrihhnthgvlhdrtghomhdpihhnvghtpeejuddrvddthedrudekuddrhedtpdhmrghilhhfrhhomhepjhhonhgrthhhrghnrdguvghrrhhitghksehlihhnuhigrdguvghvpdhnsggprhgtphhtthhopeejpdhrtghpthhtohepshhonhhgsehkvghrnhgvlhdrohhrghdprhgtphhtthhopehlihhnuhigqdhrrghiugesvhhgvghrrdhkvghrnhgvlhdrohhrghdprhgtphhtthhopeignhhisehrvgguhhgrthdrtghomhdprhgtphhtthhopehhtghhsehinhhfrhgruggvrggurdhorhhgpdhrtghpthhtohepphhmvghnii gvlhesmhholhhgvghnrdhmphhgrdguvgdprhgtphhtthhopehsuhhshhhmrgdrkhgrlhgrkhhothgrsehinhhtvghlrdgtohhm X-Xfinity-VMeta: sc=-100.00;st=legit From: Jonathan Derrick To: Song Liu , Cc: Xiao Ni , Christoph Hellwig , Paul Menzel , Sushma Kalakota , Jon Derrick Subject: [PATCH v3 3/3] md: Use optimal I/O size for last bitmap page Date: Wed, 22 Feb 2023 14:58:28 -0700 Message-Id: <20230222215828.225-4-jonathan.derrick@linux.dev> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20230222215828.225-1-jonathan.derrick@linux.dev> References: <20230222215828.225-1-jonathan.derrick@linux.dev> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-raid@vger.kernel.org From: Jon Derrick If the bitmap space has enough room, size the I/O for the last bitmap page write to the optimal I/O size for the storage device. The expanded write is checked that it won't overrun the data or metadata. The drive this was tested against has higher latencies when there are sub-4k writes due to device-side read-mod-writes of its atomic 4k write unit. This change helps increase performance by sizing the last bitmap page I/O for the device's preferred write unit, if it is given. Example Intel/Solidigm P5520 Raid10, Chunk-size 64M, bitmap-size 57228 bits $ mdadm --create /dev/md0 --level=10 --raid-devices=4 /dev/nvme{0,1,2,3}n1 --assume-clean --bitmap=internal --bitmap-chunk=64M $ fio --name=test --direct=1 --filename=/dev/md0 --rw=randwrite --bs=4k --runtime=60 Without patch: write: IOPS=1676, BW=6708KiB/s (6869kB/s)(393MiB/60001msec); 0 zone resets With patch: write: IOPS=15.7k, BW=61.4MiB/s (64.4MB/s)(3683MiB/60001msec); 0 zone resets Biosnoop: Without patch: Time Process PID Device LBA Size Lat 1.410377 md0_raid10 6900 nvme0n1 W 16 4096 0.02 1.410387 md0_raid10 6900 nvme2n1 W 16 4096 0.02 1.410374 md0_raid10 6900 nvme3n1 W 16 4096 0.01 1.410381 md0_raid10 6900 nvme1n1 W 16 4096 0.02 1.410411 md0_raid10 6900 nvme1n1 W 115346512 4096 0.01 1.410418 md0_raid10 6900 nvme0n1 W 115346512 4096 0.02 1.410915 md0_raid10 6900 nvme2n1 W 24 3584 0.43 <-- 1.410935 md0_raid10 6900 nvme3n1 W 24 3584 0.45 <-- 1.411124 md0_raid10 6900 nvme1n1 W 24 3584 0.64 <-- 1.411147 md0_raid10 6900 nvme0n1 W 24 3584 0.66 <-- 1.411176 md0_raid10 6900 nvme3n1 W 2019022184 4096 0.01 1.411189 md0_raid10 6900 nvme2n1 W 2019022184 4096 0.02 With patch: Time Process PID Device LBA Size Lat 5.747193 md0_raid10 727 nvme0n1 W 16 4096 0.01 5.747192 md0_raid10 727 nvme1n1 W 16 4096 0.02 5.747195 md0_raid10 727 nvme3n1 W 16 4096 0.01 5.747202 md0_raid10 727 nvme2n1 W 16 4096 0.02 5.747229 md0_raid10 727 nvme3n1 W 1196223704 4096 0.02 5.747224 md0_raid10 727 nvme0n1 W 1196223704 4096 0.01 5.747279 md0_raid10 727 nvme0n1 W 24 4096 0.01 <-- 5.747279 md0_raid10 727 nvme1n1 W 24 4096 0.02 <-- 5.747284 md0_raid10 727 nvme3n1 W 24 4096 0.02 <-- 5.747291 md0_raid10 727 nvme2n1 W 24 4096 0.02 <-- 5.747314 md0_raid10 727 nvme2n1 W 2234636712 4096 0.01 5.747317 md0_raid10 727 nvme1n1 W 2234636712 4096 0.02 Signed-off-by: Jon Derrick Reviewed-by: Christoph Hellwig --- drivers/md/md-bitmap.c | 33 +++++++++++++++++++++++++++++---- 1 file changed, 29 insertions(+), 4 deletions(-) diff --git a/drivers/md/md-bitmap.c b/drivers/md/md-bitmap.c index 11f4453775ee..196984fc3776 100644 --- a/drivers/md/md-bitmap.c +++ b/drivers/md/md-bitmap.c @@ -209,6 +209,28 @@ static struct md_rdev *next_active_rdev(struct md_rdev *rdev, struct mddev *mdde return NULL; } +static unsigned int optimal_io_size(struct block_device *bdev, + unsigned int last_page_size, + unsigned int io_size) +{ + if (bdev_io_opt(bdev) > bdev_logical_block_size(bdev)) + return roundup(last_page_size, bdev_io_opt(bdev)); + return io_size; +} + +static unsigned int bitmap_io_size(unsigned int io_size, unsigned int opt_size, + sector_t start, sector_t boundary) +{ + if (io_size != opt_size && + start + opt_size / SECTOR_SIZE <= boundary) + return opt_size; + else if (start + io_size / SECTOR_SIZE <= boundary) + return io_size; + + /* Overflows boundary */ + return 0; +} + static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap, struct page *page) { @@ -218,14 +240,15 @@ static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap, sector_t offset = mddev->bitmap_info.offset; sector_t ps, sboff, doff; unsigned int size = PAGE_SIZE; + unsigned int opt_size = PAGE_SIZE; bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev; if (page->index == store->file_pages - 1) { unsigned int last_page_size = store->bytes & (PAGE_SIZE - 1); if (last_page_size == 0) last_page_size = PAGE_SIZE; - size = roundup(last_page_size, - bdev_logical_block_size(bdev)); + size = roundup(last_page_size, bdev_logical_block_size(bdev)); + opt_size = optimal_io_size(bdev, last_page_size, size); } ps = page->index * (PAGE_SIZE / SECTOR_SIZE); @@ -240,7 +263,8 @@ static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap, return -EINVAL; } else if (offset < 0) { /* DATA BITMAP METADATA */ - if (offset + ps + size / SECTOR_SIZE > 0) + size = bitmap_io_size(size, opt_size, offset + ps, 0); + if (size == 0) /* bitmap runs in to metadata */ return -EINVAL; @@ -249,7 +273,8 @@ static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap, return -EINVAL; } else if (rdev->sb_start < rdev->data_offset) { /* METADATA BITMAP DATA */ - if (sboff + ps + size / SECTOR_SIZE > doff) + size = bitmap_io_size(size, opt_size, sboff + ps, doff); + if (size == 0) /* bitmap runs in to data */ return -EINVAL; } else {