linux-kernel - Re: [PATCH v3 2/4] staging: rtl8723bs: replace rtw

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <aXqHbhkXPiZVpqcS@smile.fi.intel.com>
Date: Thu, 29 Jan 2026 00:02:22 +0200
From: Andy Shevchenko <andriy.shevchenko@...el.com>
To: Minu Jin <s9430939@...er.com>
Cc: gregkh@...uxfoundation.org, hansg@...nel.org, dan.carpenter@...aro.org,
	trohan2000@...il.com, andy@...nel.org,
	linux-staging@...ts.linux.dev, linux-kernel@...r.kernel.org,
	straube.linux@...il.com
Subject: Re: [PATCH v3 2/4] staging: rtl8723bs: replace rtw_zmalloc with
 kzalloc

On Wed, Jan 28, 2026 at 11:09:52PM +0900, Minu Jin wrote:
> replaces the wrapper function rtw_zmalloc with kzalloc.
> 
> I reviewed all the call sites to determine the appropriate GFP flags:
> - GFP_KERNEL: Used for initialization and configuration paths.
>   init/probe, config/setup
> 
> - GFP_ATOMIC: Used for critical real-time paths.
>   including interrupts, timer handler, region where spinlocks are held

Below I suggest more changes, but probably they needs to be done as separate
patches. It depends on Greg's preferences.

The change itself LGTM.

...

> -	ph2c = rtw_zmalloc(sizeof(struct cmd_obj));
> +	ph2c = kzalloc(sizeof(struct cmd_obj), GFP_KERNEL);

Since you are touching this line it might be worth to check and convert to use
sizeof(*) instead (but it needs to be done with a carefulness.

>  	if (!ph2c) {
>  		res = _FAIL;
>  		goto exit;
>  	}
>  

...

> -	pcmdpriv->cmd_allocated_buf = rtw_zmalloc(MAX_CMDSZ + CMDBUFF_ALIGN_SZ);
> +	pcmdpriv->cmd_allocated_buf = kzalloc(MAX_CMDSZ + CMDBUFF_ALIGN_SZ, GFP_ATOMIC);

>  

You can drop this blank line at the same time, we usually don't split
allocations and checks.

>  	if (!pcmdpriv->cmd_allocated_buf)
>  		return -ENOMEM;
>  
>  	pcmdpriv->cmd_buf = PTR_ALIGN(pcmdpriv->cmd_allocated_buf, CMDBUFF_ALIGN_SZ);
>  
> -	pcmdpriv->rsp_allocated_buf = rtw_zmalloc(MAX_RSPSZ + 4);
> +	pcmdpriv->rsp_allocated_buf = kzalloc(MAX_RSPSZ + 4, GFP_ATOMIC);
>  

Ditto.

>  	if (!pcmdpriv->rsp_allocated_buf) {
>  		kfree(pcmdpriv->cmd_allocated_buf);

...

> -				passoc_req = rtw_zmalloc(psta->assoc_req_len);
> +				passoc_req = kzalloc(psta->assoc_req_len, GFP_ATOMIC);

>  				if (passoc_req) {
>  					assoc_req_len = psta->assoc_req_len;
>  					memcpy(passoc_req, psta->passoc_req, assoc_req_len);

This  might be a candidate to kmemdup(), but double check it (it looks like
_len:s maybe different.

...

> -	spt_band = rtw_zmalloc(sizeof(struct ieee80211_supported_band) +
> +	spt_band = kzalloc(sizeof(struct ieee80211_supported_band) +
>  			       sizeof(struct ieee80211_channel) * n_channels +
> -			       sizeof(struct ieee80211_rate) * n_bitrates);
> +			       sizeof(struct ieee80211_rate) * n_bitrates, GFP_KERNEL);

This needs a bit of array_size(), struct_size() conversion.

-- 
With Best Regards,
Andy Shevchenko