[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <edd223sc52jmxr255onvp5qwbwoyjv5douhrwxhor5qk7m2q6c@r4uqvboctxjx>
Date: Thu, 18 Dec 2025 22:38:04 +0800
From: Heming Zhao <heming.zhao@...e.com>
To: Dan Carpenter <dan.carpenter@...aro.org>
Cc: Mark Fasheh <mark@...heh.com>, Joel Becker <jlbec@...lplan.org>,
Joseph Qi <joseph.qi@...ux.alibaba.com>, Andrew Morton <akpm@...ux-foundation.org>,
Su Yue <glass.su@...e.com>, ocfs2-devel@...ts.linux.dev, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/1] ocfs2: fix uninitialized variable in
ocfs2_test_suballoc_bit()
On Thu, Dec 18, 2025 at 10:48:12AM +0300, Dan Carpenter wrote:
> The ocfs2_read_hint_group_descriptor() function sets "*released = 1;"
> but nothing ever sets it to zero so it's used without being
> initialized. Initialize it to zero.
>
> Fixes: acb8a8335a64 ("ocfs2: detect released suballocator BG for fh_to_[dentry|parent]")
> Signed-off-by: Dan Carpenter <dan.carpenter@...aro.org>
Many thanks for catching this bug.
I will try the config CONFIG_INIT_STACK_ALL_PATTERN=y
Reviewed-by: Heming Zhao <heming.zhao@...e.com>
> ---
> These days everyone has CONFIG_INIT_STACK_ALL_ZERO enabled so we never
> detect these bugs. There are a couple distros which don't, I've heard.
> But in testing, I always encourage people to use
> CONFIG_INIT_STACK_ALL_PATTERN=y
>
> fs/ocfs2/suballoc.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c
> index 72c7559c1a2b..79d1325b2111 100644
> --- a/fs/ocfs2/suballoc.c
> +++ b/fs/ocfs2/suballoc.c
> @@ -3163,7 +3163,7 @@ static int ocfs2_test_suballoc_bit(struct ocfs2_super *osb,
> struct ocfs2_group_desc *group;
> struct buffer_head *group_bh = NULL;
> u64 bg_blkno;
> - int status, quiet = 0, released;
> + int status, quiet = 0, released = 0;
>
> trace_ocfs2_test_suballoc_bit((unsigned long long)blkno,
> (unsigned int)bit);
> --
> 2.51.0
>
Powered by blists - more mailing lists