Received-SPF: pass (google.com: domain of linux-nfs-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Subject: Re: [PATCH 2/7] ceph: use locks_inode_context helper
To:     Jeff Layton <jlayton@kernel.org>, linux-fsdevel@vger.kernel.org
Cc:     linux-nfs@vger.kernel.org, ceph-devel@vger.kernel.org,
        linux-cifs@vger.kernel.org, chuck.lever@oracle.com,
        viro@zeniv.linux.org.uk, hch@lst.de
References: <20221116151726.129217-1-jlayton@kernel.org>
 <20221116151726.129217-3-jlayton@kernel.org>
From:   Xiubo Li <xiubli@redhat.com>
Message-ID: <589a0fcc-569f-e5b2-0877-c1639736ae5e@redhat.com>
Date:   Thu, 17 Nov 2022 12:56:58 +0800
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.10.1
MIME-Version: 1.0
In-Reply-To: <20221116151726.129217-3-jlayton@kernel.org>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Content-Language: en-US
Precedence: bulk


On 16/11/2022 23:17, Jeff Layton wrote:
> ceph currently doesn't access i_flctx safely. This requires a
> smp_load_acquire, as the pointer is set via cmpxchg (a release
> operation).
>
> Cc: Xiubo Li <xiubli@redhat.com>
> Signed-off-by: Jeff Layton <jlayton@kernel.org>
> ---
>   fs/ceph/locks.c | 4 ++--
>   1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/fs/ceph/locks.c b/fs/ceph/locks.c
> index 3e2843e86e27..f3b461c708a8 100644
> --- a/fs/ceph/locks.c
> +++ b/fs/ceph/locks.c
> @@ -364,7 +364,7 @@ void ceph_count_locks(struct inode *inode, int *fcntl_count, int *flock_count)
>   	*fcntl_count = 0;
>   	*flock_count = 0;
>   
> -	ctx = inode->i_flctx;
> +	ctx = locks_inode_context(inode);
>   	if (ctx) {
>   		spin_lock(&ctx->flc_lock);
>   		list_for_each_entry(lock, &ctx->flc_posix, fl_list)
> @@ -418,7 +418,7 @@ int ceph_encode_locks_to_buffer(struct inode *inode,
>   				int num_fcntl_locks, int num_flock_locks)
>   {
>   	struct file_lock *lock;
> -	struct file_lock_context *ctx = inode->i_flctx;
> +	struct file_lock_context *ctx = locks_inode_context(inode);
>   	int err = 0;
>   	int seen_fcntl = 0;
>   	int seen_flock = 0;

Thanks Jeff!

Reviewed-by: Xiubo Li <xiubli@redhat.com>