On Thu, 2018-07-26 at 22:22 +0800, Coly Li wrote:
> Global variable bcache_debug is firstly initialized in bch_debug_init(),
> and destroyed in bch_debug_exit(). bch_debug_init() is called in
> bcache_init() with many other functions, if one of the previous calling
> onces failed, bcache_exit() will be called in the failure path.
> 
> The problem is, if bcache_init() fails before bch_debug_init() is called,
> then in bcache_exit() when bch_debug_exit() is called to destroy global
> variable bcache_debug, at this moment bcache_debug is unndefined, then the
> test of "if (!IS_ERR_OR_NULL(bcache_debug))" might be buggy.
> 
> This patch initializes global varabile bcache_debug to be NULL, to make
> the failure code path to be predictable.
> 
> Signed-off-by: Coly Li <[email protected]>
> ---
>  drivers/md/bcache/debug.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/md/bcache/debug.c b/drivers/md/bcache/debug.c
> index 57f8f5aeee55..24b0eb65ddec 100644
> --- a/drivers/md/bcache/debug.c
> +++ b/drivers/md/bcache/debug.c
> @@ -17,7 +17,7 @@
>  #include <linux/random.h>
>  #include <linux/seq_file.h>
>  
> -struct dentry *bcache_debug;
> +struct dentry *bcache_debug = NULL;
>  
>  #ifdef CONFIG_BCACHE_DEBUG

Please verify patches with checkpatch before posting these. Checkpatch is namely
able to detect that this patch is useless:

$ scripts/checkpatch.pl 
\[PATCH_8_9\]_bcache\:_initiate_bcache_debug_to_NULL.mbox 
ERROR: do not initialise globals to NULL
#217: FILE: drivers/md/bcache/debug.c:20:
+struct dentry *bcache_debug = NULL;

Bart.

Reply via email to