[PATCH] erofs: fix inline pcluster memory leak

[Yue Hu]

On Fri,  2 Dec 2022 11:33:27 +0800
Gao Xiang <hsiangkao at linux.alibaba.com> wrote:

> Inline pcluster should be freed right after it is decompressed.
> Otherwise inline pclusters will be leaked.
> 
> Reported-by: syzbot+6f8cd9a0155b366d227f at syzkaller.appspotmail.com
> Fixes: cecf864d3d76 ("erofs: support inline data decompression")
> Cc: Yue Hu <huyue2 at yulong.com>
> Signed-off-by: Gao Xiang <hsiangkao at linux.alibaba.com>

Reviewed-by: Yue Hu <huyue2 at coolpad.com>

Thanks.

> ---
>  fs/erofs/zdata.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
> index ab22100be861..e14e6c32e70d 100644
> --- a/fs/erofs/zdata.c
> +++ b/fs/erofs/zdata.c
> @@ -496,7 +496,8 @@ static int z_erofs_register_pcluster(struct z_erofs_decompress_frontend *fe)
>  	struct erofs_workgroup *grp;
>  	int err;
>  
> -	if (!(map->m_flags & EROFS_MAP_ENCODED)) {
> +	if (!(map->m_flags & EROFS_MAP_ENCODED) ||
> +	    !(map->m_pa >> PAGE_SHIFT)) {
>  		DBG_BUGON(1);
>  		return -EFSCORRUPTED;
>  	}
> @@ -1114,6 +1115,8 @@ static int z_erofs_decompress_pcluster(struct z_erofs_decompress_backend *be,
>  	/* pcluster lock MUST be taken before the following line */
>  	WRITE_ONCE(pcl->next, Z_EROFS_PCLUSTER_NIL);
>  	mutex_unlock(&pcl->lock);
> +	if (z_erofs_is_inline_pcluster(pcl))
> +		z_erofs_free_pcluster(pcl);
>  	return err;
>  }
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ozlabs.org/pipermail/linux-erofs/attachments/20221202/3d3ec5c4/attachment.htm>