Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH] btrfs: relocation: fix misplaced barrier in
 reloc_root_is_dead
To:     Baptiste Lepers <baptiste.lepers@gmail.com>
Cc:     Chris Mason <clm@fb.com>, Josef Bacik <josef@toxicpanda.com>,
        David Sterba <dsterba@suse.com>, Qu Wenruo <wqu@suse.com>,
        linux-btrfs@vger.kernel.org, linux-kernel@vger.kernel.org
References: <20210526060947.26159-1-baptiste.lepers@gmail.com>
From:   Qu Wenruo <quwenruo.btrfs@gmx.com>
Message-ID: <9fcd47ac-2642-88be-1bf0-7b920baefb10@gmx.com>
Date:   Wed, 26 May 2021 14:13:25 +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: <20210526060947.26159-1-baptiste.lepers@gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: quoted-printable
Precedence: bulk


On 2021/5/26 =E4=B8=8B=E5=8D=882:09, Baptiste Lepers wrote:
> Fix a misplaced barrier in reloc_root_is_dead. The
> BTRFS_ROOT_DEAD_RELOC_TREE bit should be checked before accessing
> reloc_root.
>
> The fix forces the order documented in the original commit:
> "In the cross section C-D, either caller gets the DEAD bit set, avoiding
> access reloc_root no matter if it's safe or not.  Or caller get the DEAD
> bit cleared, then access reloc_root, which is already NULL, nothing will
> be wrong."
>
> static bool reloc_root_is_dead()
>      smp_rmb(); <--- misplaced
>      if (test_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state))
>            return true;
>      return false;
> }

Exactly what I originally purposed to David, we didn't reach an
agreement on where the barrier should be.

At least I'm completely fine with this patch.

Reviewed-by: Qu Wenruo <wqu@suse.com>

Thanks,
Qu
>
> static bool have_reloc_root(struct btrfs_root *root)
> {
>         if (reloc_root_is_dead(root))
>                 return false;
>         <--- the barrier should happen here
>         if (!root->reloc_root)
>                 return false;
>         return true;
> }
>
> Fixes: 6282675e6708e ("btrfs: relocation: fix reloc_root lifespan and ac=
cess")
> Signed-off-by: Baptiste Lepers <baptiste.lepers@gmail.com>
> ---
>   fs/btrfs/relocation.c | 5 ++---
>   1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
> index b70be2ac2e9e..8cddcce56bf4 100644
> --- a/fs/btrfs/relocation.c
> +++ b/fs/btrfs/relocation.c
> @@ -289,15 +289,14 @@ static int update_backref_cache(struct btrfs_trans=
_handle *trans,
>
>   static bool reloc_root_is_dead(struct btrfs_root *root)
>   {
> +	bool is_dead =3D test_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state);
>   	/*
>   	 * Pair with set_bit/clear_bit in clean_dirty_subvols and
>   	 * btrfs_update_reloc_root. We need to see the updated bit before
>   	 * trying to access reloc_root
>   	 */
>   	smp_rmb();
> -	if (test_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state))
> -		return true;
> -	return false;
> +	return is_dead;
>   }
>
>   /*
>