[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <448062a9-43f8-c7ad-0b4d-ddf892713d9e@huaweicloud.com>
Date: Mon, 6 May 2024 22:42:02 +0800
From: Zhang Yi <yi.zhang@...weicloud.com>
To: Kemeng Shi <shikemeng@...weicloud.com>
Cc: linux-ext4@...r.kernel.org, linux-kernel@...r.kernel.org, tytso@....edu,
jack@...e.com
Subject: Re: [PATCH 1/9] jbd2: avoid memleak in
jbd2_journal_write_metadata_buffer
On 2024/5/6 22:17, Kemeng Shi wrote:
> The new_bh is from alloc_buffer_head, we should call free_buffer_head to
> free it in error case.
Oh, yeah, this is a separate bh, so it should be freed explicitly, it
looks good to me.
Reviewed-by: Zhang Yi <yi.zhang@...wei.com>
>
> Signed-off-by: Kemeng Shi <shikemeng@...weicloud.com>
> ---
> fs/jbd2/journal.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
> index b6c114c11b97..207b24e12ce9 100644
> --- a/fs/jbd2/journal.c
> +++ b/fs/jbd2/journal.c
> @@ -399,6 +399,7 @@ int jbd2_journal_write_metadata_buffer(transaction_t *transaction,
> tmp = jbd2_alloc(bh_in->b_size, GFP_NOFS);
> if (!tmp) {
> brelse(new_bh);
> + free_buffer_head(new_bh);
> return -ENOMEM;
> }
> spin_lock(&jh_in->b_state_lock);
>
Powered by blists - more mailing lists