From: Chao Yu <yuchao0@huawei.com> To: Jaegeuk Kim <jaegeuk@kernel.org>, Sayali Lokhande <sayalil@codeaurora.org> Cc: <linux-f2fs-devel@lists.sourceforge.net>, <stummala@codeaurora.org>, <linux-kernel@vger.kernel.org> Subject: Re: [PATCH V4] f2fs: Avoid double lock for cp_rwsem during checkpoint Date: Sat, 9 May 2020 11:02:25 +0800 [thread overview] Message-ID: <0902037e-998d-812e-53e7-90ea7b9957eb@huawei.com> (raw) In-Reply-To: <20200508161052.GA49579@google.com> On 2020/5/9 0:10, Jaegeuk Kim wrote: > Hi Sayali, > > In order to address the perf regression, how about this? > >>From 48418af635884803ffb35972df7958a2e6649322 Mon Sep 17 00:00:00 2001 > From: Jaegeuk Kim <jaegeuk@kernel.org> > Date: Fri, 8 May 2020 09:08:37 -0700 > Subject: [PATCH] f2fs: avoid double lock for cp_rwsem during checkpoint > > There could be a scenario where f2fs_sync_node_pages gets > called during checkpoint, which in turn tries to flush > inline data and calls iput(). This results in deadlock as > iput() tries to hold cp_rwsem, which is already held at the > beginning by checkpoint->block_operations(). > > Call stack : > > Thread A Thread B > f2fs_write_checkpoint() > - block_operations(sbi) > - f2fs_lock_all(sbi); > - down_write(&sbi->cp_rwsem); > > - open() > - igrab() > - write() write inline data > - unlink() > - f2fs_sync_node_pages() > - if (is_inline_node(page)) > - flush_inline_data() > - ilookup() > page = f2fs_pagecache_get_page() > if (!page) > goto iput_out; > iput_out: > -close() > -iput() > iput(inode); > - f2fs_evict_inode() > - f2fs_truncate_blocks() > - f2fs_lock_op() > - down_read(&sbi->cp_rwsem); > > Fixes: 2049d4fcb057 ("f2fs: avoid multiple node page writes due to inline_data") > Signed-off-by: Sayali Lokhande <sayalil@codeaurora.org> > Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org> > --- > fs/f2fs/node.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c > index 1db8cabf727ef..626d7daca09de 100644 > --- a/fs/f2fs/node.c > +++ b/fs/f2fs/node.c > @@ -1870,8 +1870,8 @@ int f2fs_sync_node_pages(struct f2fs_sb_info *sbi, > goto continue_unlock; > } > > - /* flush inline_data */ > - if (is_inline_node(page)) { > + /* flush inline_data, if it's not sync path. */ > + if (do_balance && is_inline_node(page)) { IIRC, this flow was designed to avoid running out of free space issue during checkpoint: 2049d4fcb057 ("f2fs: avoid multiple node page writes due to inline_data") The sceanrio is: 1. create fully node blocks 2. flush node blocks 3. write inline_data for all the node blocks again 4. flush node blocks redundantly I guess this may cause failing one case of fstest. Since block_operations->f2fs_sync_inode_meta has synced inode cache to inode page, so in block_operations->f2fs_sync_node_pages, could we check nlink before flush_inline_data(): if (is_inline_node(page)) { if (IS_INODE(page) && raw_inode_page->i_links) { flush_inline_data() } } > clear_inline_node(page); > unlock_page(page); > flush_inline_data(sbi, ino_of_node(page)); >
WARNING: multiple messages have this Message-ID (diff)
From: Chao Yu <yuchao0@huawei.com> To: Jaegeuk Kim <jaegeuk@kernel.org>, Sayali Lokhande <sayalil@codeaurora.org> Cc: linux-kernel@vger.kernel.org, linux-f2fs-devel@lists.sourceforge.net Subject: Re: [f2fs-dev] [PATCH V4] f2fs: Avoid double lock for cp_rwsem during checkpoint Date: Sat, 9 May 2020 11:02:25 +0800 [thread overview] Message-ID: <0902037e-998d-812e-53e7-90ea7b9957eb@huawei.com> (raw) In-Reply-To: <20200508161052.GA49579@google.com> On 2020/5/9 0:10, Jaegeuk Kim wrote: > Hi Sayali, > > In order to address the perf regression, how about this? > >>From 48418af635884803ffb35972df7958a2e6649322 Mon Sep 17 00:00:00 2001 > From: Jaegeuk Kim <jaegeuk@kernel.org> > Date: Fri, 8 May 2020 09:08:37 -0700 > Subject: [PATCH] f2fs: avoid double lock for cp_rwsem during checkpoint > > There could be a scenario where f2fs_sync_node_pages gets > called during checkpoint, which in turn tries to flush > inline data and calls iput(). This results in deadlock as > iput() tries to hold cp_rwsem, which is already held at the > beginning by checkpoint->block_operations(). > > Call stack : > > Thread A Thread B > f2fs_write_checkpoint() > - block_operations(sbi) > - f2fs_lock_all(sbi); > - down_write(&sbi->cp_rwsem); > > - open() > - igrab() > - write() write inline data > - unlink() > - f2fs_sync_node_pages() > - if (is_inline_node(page)) > - flush_inline_data() > - ilookup() > page = f2fs_pagecache_get_page() > if (!page) > goto iput_out; > iput_out: > -close() > -iput() > iput(inode); > - f2fs_evict_inode() > - f2fs_truncate_blocks() > - f2fs_lock_op() > - down_read(&sbi->cp_rwsem); > > Fixes: 2049d4fcb057 ("f2fs: avoid multiple node page writes due to inline_data") > Signed-off-by: Sayali Lokhande <sayalil@codeaurora.org> > Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org> > --- > fs/f2fs/node.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c > index 1db8cabf727ef..626d7daca09de 100644 > --- a/fs/f2fs/node.c > +++ b/fs/f2fs/node.c > @@ -1870,8 +1870,8 @@ int f2fs_sync_node_pages(struct f2fs_sb_info *sbi, > goto continue_unlock; > } > > - /* flush inline_data */ > - if (is_inline_node(page)) { > + /* flush inline_data, if it's not sync path. */ > + if (do_balance && is_inline_node(page)) { IIRC, this flow was designed to avoid running out of free space issue during checkpoint: 2049d4fcb057 ("f2fs: avoid multiple node page writes due to inline_data") The sceanrio is: 1. create fully node blocks 2. flush node blocks 3. write inline_data for all the node blocks again 4. flush node blocks redundantly I guess this may cause failing one case of fstest. Since block_operations->f2fs_sync_inode_meta has synced inode cache to inode page, so in block_operations->f2fs_sync_node_pages, could we check nlink before flush_inline_data(): if (is_inline_node(page)) { if (IS_INODE(page) && raw_inode_page->i_links) { flush_inline_data() } } > clear_inline_node(page); > unlock_page(page); > flush_inline_data(sbi, ino_of_node(page)); > _______________________________________________ Linux-f2fs-devel mailing list Linux-f2fs-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
next prev parent reply other threads:[~2020-05-09 3:02 UTC|newest] Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top 2020-04-30 10:58 [PATCH V4] f2fs: Avoid double lock for cp_rwsem during checkpoint Sayali Lokhande 2020-04-30 10:58 ` [f2fs-dev] " Sayali Lokhande 2020-05-06 6:21 ` Chao Yu 2020-05-06 6:21 ` [f2fs-dev] " Chao Yu 2020-05-08 8:58 ` [f2fs] da55a3ad36: filebench.sum_bytes_mb/s -44.4% regression kernel test robot 2020-05-08 8:58 ` kernel test robot 2020-05-08 16:10 ` [PATCH V4] f2fs: Avoid double lock for cp_rwsem during checkpoint Jaegeuk Kim 2020-05-08 16:10 ` [f2fs-dev] " Jaegeuk Kim 2020-05-09 3:02 ` Chao Yu [this message] 2020-05-09 3:02 ` Chao Yu 2020-05-09 19:03 ` Jaegeuk Kim 2020-05-09 19:03 ` [f2fs-dev] " Jaegeuk Kim 2020-05-11 1:28 ` Chao Yu 2020-05-11 1:28 ` [f2fs-dev] " Chao Yu 2020-05-11 22:11 ` Jaegeuk Kim 2020-05-11 22:11 ` [f2fs-dev] " Jaegeuk Kim 2020-05-12 1:57 ` Chao Yu 2020-05-12 1:57 ` [f2fs-dev] " Chao Yu 2020-05-12 3:24 ` Jaegeuk Kim 2020-05-12 3:24 ` [f2fs-dev] " Jaegeuk Kim 2020-05-12 6:49 ` Chao Yu 2020-05-12 6:49 ` [f2fs-dev] " Chao Yu
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=0902037e-998d-812e-53e7-90ea7b9957eb@huawei.com \ --to=yuchao0@huawei.com \ --cc=jaegeuk@kernel.org \ --cc=linux-f2fs-devel@lists.sourceforge.net \ --cc=linux-kernel@vger.kernel.org \ --cc=sayalil@codeaurora.org \ --cc=stummala@codeaurora.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.