Re: [PATCH 4/5] staging: rtl8723bs: replace function with error handling alternative
From: Dan Carpenter
Date: Wed Apr 01 2026 - 04:46:41 EST
On Tue, Mar 31, 2026 at 05:32:53PM +0200, Omer El Idrissi wrote:
> Replace the use of rtw_set_hal_ops with rtw_hal_data_init in
> rtw_sdio_if1_init , which actually returns error or success and not
> void.
> rtw_set_hal_ops literally only calls rtw_hal_data_init and just ignores the
> possibility of errors.
>
This is a behavior change and it's quite dangerous. A lot of code only
works because there is no error handling. We can't merge it without
testing unless it causes a security issue or something. For example,
not checking the results of allocations could cause a crash so maybe
that's a security bug.
> @@ -250,8 +244,9 @@ static struct adapter *rtw_sdio_if1_init(struct dvobj_priv *dvobj, const struct
> /* 3 3. init driver special setting, interface, OS and hardware relative */
>
> /* 4 3.1 set hardware operation functions */
> - rtw_set_hal_ops(padapter);
> -
> + /* allocates padapter->HalData */
> + if (rtw_hal_data_init(padapter))
> + goto free_adapter;
I don't want to see checks in this style. I always want them to be
in this format:
ret = rtw_hal_data_init(padapter);
if (ret)
...
But in this case, just leave it because adding a new check would have
to be tested.
regards,
dan carpenter
>
> /* 3 5. initialize Chip version */
> padapter->intf_start = &sd_intf_start;
> --
> 2.51.0
>