Received-SPF: pass (google.com: domain of linux-kernel+bounces-195826-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Date: Thu, 30 May 2024 22:28:34 +0300
From: Dan Carpenter <dan.carpenter@linaro.org>
To: Thorsten Blum <thorsten.blum@toblux.com>
Cc: markus.elfring@web.de, Shyam-sundar.S-k@amd.com, hdegoede@redhat.com,
	ilpo.jarvinen@linux.intel.com, kernel-janitors@vger.kernel.org,
	linux-kernel@vger.kernel.org, platform-driver-x86@vger.kernel.org
Subject: Re: [PATCH v2] platform/x86/amd/pmf: Use memdup_user()
Message-ID: <1512f46e-112c-47ef-8620-add0f04ce86d@moroto.mountain>
References: <50d10f0b-d108-4057-be9c-a90e2398e810@web.de>
 <20240527083628.210491-2-thorsten.blum@toblux.com>
 <4c2c8a76-17c7-4dbb-96ff-8488c8e953ff@moroto.mountain>
 <1D5BDE66-ABB3-41D9-ACB0-FBAC207D55DC@toblux.com>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1D5BDE66-ABB3-41D9-ACB0-FBAC207D55DC@toblux.com>

On Thu, May 30, 2024 at 04:15:51PM +0200, Thorsten Blum wrote:
> Hi Dan,
> 
> On 27. May 2024, at 12:38, Dan Carpenter <dan.carpenter@linaro.org> wrote:
> > Also this check isn't great:
> > 
> > if (dev->policy_sz < header->length + 512)
> > 
> > header->length is a u32 that comes from the user, so the addition can
> > overflow.  I can't immediately see how to exploit this though since we
> > don't seem to use header->length after this (by itself).
> 
> How about
> 
> 	if (header->length > U32_MAX - 512 || dev->policy_sz < header->length + 512)
> 		return -EINVAL;
> 
> to prevent a possible overflow?

I've been thinking about this and actually we could do something simpler:

	if (dev->policy_sz < size_add(header->length, 512)) {

> 
> header->length is used in the next line
> 
> 	dev->policy_sz = header->length + 512;

Yeah, but it's not used by itself.  The "header->length + 512" has been
verified as a valid value whether it overflows or not.  Only
"header->length" is wrong.

> 
> and if the addition overflows, we end up setting dev->policy_sz to an 
> invalid value.

regards,
dan carpenter