linux-ext4 - [PATCH 08/35] fs: Fix performance regression in clean_bdev

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu,  1 Jun 2017 11:32:18 +0200
From:   Jan Kara <jack@...e.cz>
To:     <linux-mm@...ck.org>
Cc:     Hugh Dickins <hughd@...gle.com>,
        David Howells <dhowells@...hat.com>,
        linux-afs@...ts.infradead.org,
        Ryusuke Konishi <konishi.ryusuke@....ntt.co.jp>,
        linux-nilfs@...r.kernel.org, Bob Peterson <rpeterso@...hat.com>,
        cluster-devel@...hat.com, Jaegeuk Kim <jaegeuk@...nel.org>,
        linux-f2fs-devel@...ts.sourceforge.net, tytso@....edu,
        linux-ext4@...r.kernel.org, Ilya Dryomov <idryomov@...il.com>,
        "Yan, Zheng" <zyan@...hat.com>, ceph-devel@...r.kernel.org,
        linux-btrfs@...r.kernel.org, David Sterba <dsterba@...e.com>,
        "Darrick J . Wong" <darrick.wong@...cle.com>,
        linux-xfs@...r.kernel.org,
        Nadia Yvette Chambers <nyc@...omorphy.com>,
        Jan Kara <jack@...e.cz>
Subject: [PATCH 08/35] fs: Fix performance regression in clean_bdev_aliases()

Commit e64855c6cfaa "fs: Add helper to clean bdev aliases under a bh and
use it" added a wrapper for clean_bdev_aliases() that invalidates bdev
aliases underlying a single buffer head. However this has caused a
performance regression for bonnie++ benchmark on ext4 filesystem when
delayed allocation is turned off (ext3 mode) - average of 3 runs:

Hmean SeqOut Char  164787.55 (  0.00%) 107189.06 (-34.95%)
Hmean SeqOut Block 219883.89 (  0.00%) 168870.32 (-23.20%)

The reason for this regression is that clean_bdev_aliases() is slower
when called for a single block because pagevec_lookup() it uses will end
up iterating through the radix tree until it finds a page (which may
take a while) but we are only interested whether there's a page at a
particular index.

Fix the problem by using pagevec_lookup_range() instead which avoids the
needless iteration.

Fixes: e64855c6cfaa0a80c1b71c5f647cb792dc436668
Signed-off-by: Jan Kara <jack@...e.cz>
---
 fs/buffer.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/fs/buffer.c b/fs/buffer.c
index fe0ee01c5a44..d63b22e50f38 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -1632,19 +1632,18 @@ void clean_bdev_aliases(struct block_device *bdev, sector_t block, sector_t len)
 	struct pagevec pvec;
 	pgoff_t index = block >> (PAGE_SHIFT - bd_inode->i_blkbits);
 	pgoff_t end;
-	int i;
+	int i, count;
 	struct buffer_head *bh;
 	struct buffer_head *head;
 
 	end = (block + len - 1) >> (PAGE_SHIFT - bd_inode->i_blkbits);
 	pagevec_init(&pvec, 0);
-	while (index <= end && pagevec_lookup(&pvec, bd_mapping, &index,
-			min(end - index, (pgoff_t)PAGEVEC_SIZE - 1) + 1)) {
-		for (i = 0; i < pagevec_count(&pvec); i++) {
+	while (pagevec_lookup_range(&pvec, bd_mapping, &index, end,
+				    PAGEVEC_SIZE)) {
+		count = pagevec_count(&pvec);
+		for (i = 0; i < count; i++) {
 			struct page *page = pvec.pages[i];
 
-			if (page->index > end)
-				break;
 			if (!page_has_buffers(page))
 				continue;
 			/*
@@ -1674,6 +1673,9 @@ void clean_bdev_aliases(struct block_device *bdev, sector_t block, sector_t len)
 		}
 		pagevec_release(&pvec);
 		cond_resched();
+		/* End of range already reached? */
+		if (index > end || !index)
+			break;
 	}
 }
 EXPORT_SYMBOL(clean_bdev_aliases);
-- 
2.12.3