Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
Date:   Fri, 6 May 2022 19:22:09 +0800
From:   Wang Cheng <wanngchenng@gmail.com>
To:     Dan Carpenter <dan.carpenter@oracle.com>
Cc:     paskripkin@gmail.com, linux-staging@lists.linux.dev,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 1/2] staging: rtl8712: fix uninit-value in usb_read8()
 and friends
Message-ID: <20220506112209.hslb67cuinntusuc@ppc.localdomain>
References: <cover.1651760402.git.wanngchenng@gmail.com>
 <b9b7a6ee02c02aa28054f5cf16129977775f3cd9.1651760402.git.wanngchenng@gmail.com>
 <20220506071023.GB4031@kadam>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220506071023.GB4031@kadam>
Precedence: bulk

On 22/05/06 10:10AM, Dan Carpenter wrote:
> Setting "data = 0" will silence the KMSAN warnings but it doesn't fix
> the bug which is that r8712_usbctrl_vendorreq() treats partial reads
> as success.
> 
> The usb_control_msg() returns negatives on total failure and it returns
> small positives on partial failure.  So take the code that I gave you
> before and put that into r8712_usbctrl_vendorreq().  That's patch 1.
> These patches become 2 and 3.

Ah, you mentioned in previous review "But then another problem is that
"status" can be less than "len"." I missed it then.

- w

> 
>         status = usb_control_msg();
>         if (status < 0)
>                 goto free;
>         if (status != len) {
>                 status = -EREMOTEIO;
>                 goto free;
>         }
>         if (requesttype == 0x01)
>                 memcpy(pdata, pIo_buf, status);
> 
> regards,
> dan carpenter
>