[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <6cfb8f00-a36c-03b9-ffa7-12fe37a19317@huawei.com>
Date: Mon, 27 Apr 2020 14:08:39 +0800
From: Chao Yu <yuchao0@...wei.com>
To: Sayali Lokhande <sayalil@...eaurora.org>, <jaegeuk@...nel.org>,
<linux-f2fs-devel@...ts.sourceforge.net>
CC: <stummala@...eaurora.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH V2] f2fs: Avoid double lock for cp_rwsem during checkpoint
On 2020/4/27 14:00, Sayali Lokhande wrote:
> 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:
>
> - iput()
^^^^
- close()
- iput()
> iput(inode);
> - f2fs_evict_inode()
> - f2fs_truncate_blocks()
> - f2fs_lock_op()
> - down_read(&sbi->cp_rwsem);
>
> Signed-off-by: Sayali Lokhande <sayalil@...eaurora.org>
Reviewed-by: Chao Yu <yuchao0@...wei.com>
Thanks,
Powered by blists - more mailing lists