linux-kernel - Re: [PATCH 3/3] dm-pcache: avoid leaking invalid metadata in pcache_meta_find

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20251103113803.00001448@huawei.com>
Date: Mon, 3 Nov 2025 11:38:03 +0000
From: Jonathan Cameron <jonathan.cameron@...wei.com>
To: Li Chen <me@...ux.beauty>
CC: <dm-devel@...ts.linux.dev>, <linux-kernel@...r.kernel.org>, Dongsheng Yang
	<dongsheng.yang@...ux.dev>, Zheng Gu <cengku@...il.com>
Subject: Re: [PATCH 3/3] dm-pcache: avoid leaking invalid metadata in
 pcache_meta_find_latest()

On Thu, 30 Oct 2025 20:33:21 +0800
Li Chen <me@...ux.beauty> wrote:

> From: Li Chen <chenl311@...natelecom.cn>
> 
> Before this change pcache_meta_find_latest() was copying each
> slot directly into meta_ret while scanning. If no valid slot
> was found and the function returned NULL, meta_ret still held
> whatever was last copied (possibly CRC-bad). Later users
> (e.g. cache_segs_init) could mistakenly trust that data.
> 
> Allocate a temporary buffer instead and only populate meta_ret after a
> valid/latest header is found. If no valid header exists we return NULL
> without touching meta_ret.
> 
> Also add __free(kvfree) so the temporary buffer is always freed, and
> include the needed headers.
> 
> Signed-off-by: Li Chen <chenl311@...natelecom.cn>
> ---
>  drivers/md/dm-pcache/pcache_internal.h | 13 +++++++++----
>  1 file changed, 9 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/md/dm-pcache/pcache_internal.h b/drivers/md/dm-pcache/pcache_internal.h
> index b7a3319d2bd3e..ac28f9dd2986f 100644
> --- a/drivers/md/dm-pcache/pcache_internal.h
> +++ b/drivers/md/dm-pcache/pcache_internal.h
> @@ -4,6 +4,8 @@
>  
>  #include <linux/delay.h>
>  #include <linux/crc32c.h>
> +#include <linux/slab.h>
> +#include <linux/cleanup.h>
>  
>  #define pcache_err(fmt, ...)							\
>  	pr_err("dm-pcache: %s:%u " fmt, __func__, __LINE__, ##__VA_ARGS__)
> @@ -79,14 +81,17 @@ static inline void __must_check *pcache_meta_find_latest(struct pcache_meta_head
>  					u32 meta_size, u32 meta_max_size,
>  					void *meta_ret)
>  {
> -	struct pcache_meta_header *meta, *latest = NULL;
> +	struct pcache_meta_header *latest = NULL;
> +	struct pcache_meta_header *meta __free(kvfree);
>  	u32 i, seq_latest = 0;
> -	void *meta_addr;
>  
> -	meta = meta_ret;
> +	meta = kvzalloc(meta_size, GFP_KERNEL);
See the guidance notes in cleanup.h  THis hsould be

	struct pcache_meta_header *meta __free(kvfree) =
		kvzalloc(meta_size, GFP_KERNEL);

That is the constructor and destructor must be together. Inline variable
declarations are fine for this one type of use.

> +	if (!meta)
> +		return ERR_PTR(-ENOMEM);
>  
>  	for (i = 0; i < PCACHE_META_INDEX_MAX; i++) {
> -		meta_addr = (void *)header + (i * meta_max_size);
> +		void *meta_addr = (void *)header + (i * meta_max_size);
> +
>  		if (copy_mc_to_kernel(meta, meta_addr, meta_size)) {
>  			pcache_err("hardware memory error when copy meta");
>  			return ERR_PTR(-EIO);