Received-SPF: pass (google.com: domain of linux-ext4-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Baokun Li <libaokun1@huawei.com>
To:     <jack@suse.cz>
CC:     <linux-fsdevel@vger.kernel.org>, <linux-ext4@vger.kernel.org>,
        <linux-kernel@vger.kernel.org>, <yi.zhang@huawei.com>,
        <yangerkun@huawei.com>, <chengzhihao1@huawei.com>,
        <yukuai3@huawei.com>, <libaokun1@huawei.com>
Subject: [PATCH v2 0/7] quota: fix race condition between dqput() and dquot_mark_dquot_dirty()
Date:   Wed, 28 Jun 2023 21:21:48 +0800
Message-ID: <20230628132155.1560425-1-libaokun1@huawei.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 7BIT
Content-Type:   text/plain; charset=US-ASCII
Precedence: bulk

Hello Honza,

This is a solution that uses dquot_srcu to avoid race condition between
dqput() and dquot_mark_dquot_dirty(). I performed a 24+h fault injection
stress test (6 VMs, 4 test threads per VM) and have not found any problems.
And I tested the performance based on the latest mainline (6aeadf7896bf),
the patch set did not lead to performance degradation, and even a little
bit of improvement.

V1->V2:
	Modify the solution to use dquot_srcu.

Baokun Li (7):
  quota: factor out dquot_write_dquot()
  quota: add new global dquot list releasing_dquots
  quota: rename dquot_active() to inode_dquot_active()
  quota: add new helper dquot_active()
  quota: fix dqput() to follow the guarantees dquot_srcu should provide
  quota: simplify drop_dquot_ref()
  quota: remove unused function put_dquot_list()

 fs/quota/dquot.c | 237 +++++++++++++++++++++++++++--------------------
 1 file changed, 134 insertions(+), 103 deletions(-)

-- 
2.31.1