Re: [PATCH 1/2] interconnect: Add character pointer initialization
From: Greg KH
Date: Fri Aug 30 2024 - 06:35:27 EST
On Fri, Aug 30, 2024 at 06:22:44PM +0800, Yibin Ding wrote:
> From: Yibin Ding <Yibin.ding@xxxxxxxxxx>
>
> When accessing a node whose data type is a character pointer and has not
> been initialized, a crash will occur due to accessing a null pointer. So
> it is necessary to add the operation of initializing the character pointer.
> Since the debugfs_write_file_str() function performs a kfree() operation
> on the node data, memory is allocated to the node pointer during
> initialization will be released when data is written to the node.
>
> Signed-off-by: Yibin Ding <Yibin.ding@xxxxxxxxxx>
> ---
> drivers/interconnect/debugfs-client.c | 10 ++++++++--
> 1 file changed, 8 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/interconnect/debugfs-client.c b/drivers/interconnect/debugfs-client.c
> index bc3fd8a7b9eb..d62ba56b7bbe 100644
> --- a/drivers/interconnect/debugfs-client.c
> +++ b/drivers/interconnect/debugfs-client.c
> @@ -16,6 +16,7 @@
> #undef INTERCONNECT_ALLOW_WRITE_DEBUGFS
>
> #if defined(INTERCONNECT_ALLOW_WRITE_DEBUGFS) && defined(CONFIG_DEBUG_FS)
> +#define INITNODE_SIZE 1
Why is this needed? Why not just use the size of the structure?
>
> static LIST_HEAD(debugfs_paths);
> static DEFINE_MUTEX(debugfs_lock);
> @@ -147,8 +148,13 @@ int icc_debugfs_client_init(struct dentry *icc_dir)
>
> client_dir = debugfs_create_dir("test_client", icc_dir);
>
> - debugfs_create_str("src_node", 0600, client_dir, &src_node);
> - debugfs_create_str("dst_node", 0600, client_dir, &dst_node);
> + src_node = kzalloc(INITNODE_SIZE, GFP_KERNEL);
> + dst_node = kzalloc(INITNODE_SIZE, GFP_KERNEL);
Wow, how did this ever work at all?
What commit id does this fix?
And where are you freeing this memory you just allocated?
thanks,
greg k-h