Re: [PATCH 1/2] staging: rtl8723bs: remove dead rtw_atimdone_event_callback()

[Dan Carpenter <error27@gmail.com>]

On Fri, May 22, 2026 at 10:02:55PM +0300, Andrei Khomenkov wrote:
> diff --git a/drivers/staging/rtl8723bs/core/rtw_mlme_ext.c b/drivers/staging/rtl8723bs/core/rtw_mlme_ext.c
> index a86d6f97cf02..c56082cd5264 100644
> --- a/drivers/staging/rtl8723bs/core/rtw_mlme_ext.c
> +++ b/drivers/staging/rtl8723bs/core/rtw_mlme_ext.c
> @@ -5619,7 +5619,7 @@ static struct fwevent wlanevents[] = {
>  	{0, &rtw_joinbss_event_callback},		/*10*/
>  	{sizeof(struct stassoc_event), &rtw_stassoc_event_callback},
>  	{sizeof(struct stadel_event), &rtw_stadel_event_callback},
> -	{0, &rtw_atimdone_event_callback},
> +	{0, rtw_dummy_event_callback},
>  	{0, rtw_dummy_event_callback},
>  	{0, NULL},	/*15*/
>  	{0, NULL},

This patch is fine...

Reviewed-by: Dan Carpenter <error27@gmail.com>

But, dummy functions are really discouraged.  It's supposed to be that
the caller checks for NULL and then doesn't call it.  No point in having
a do nothing dummy function when we could just store a NULL pointer there
instead.

And most of that table is NULL pointers!  I looked at the callers and
it's alarming that I don't immediately see any NULL checks.

  5657          /*  checking if event size match the event parm size */
  5658          if ((wlanevents[evt_code].parmsize != 0) &&

If the callback is NULL then parmsize is 0 so we don't care about the
size...  Just continue.

  5659              (wlanevents[evt_code].parmsize != evt_sz))
  5660                  goto _abort_event_;
  5661  
  5662          atomic_inc(&pevt_priv->event_seq);
  5663  
  5664          peventbuf += 2;
  5665  
  5666          if (peventbuf) {
  5667                  event_callback = wlanevents[evt_code].event_callback;

Get the call back.

  5668                  event_callback(padapter, (u8 *)peventbuf);

And crash...  I guess we are relying on this table to be in sync with
the wlancmds[] table in drivers/staging/rtl8723bs/core/rtw_cmd.c.
That seems pretty risky to me.

  5669  
  5670                  pevt_priv->evt_done_cnt++;
  5671          }

regards,
dan carpenter