Received-SPF: pass (google.com: domain of linux-kernel+bounces-190484-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Message-ID: <59d4bf96-9a94-4e2c-9636-81e79d572438@redhat.com>
Date: Mon, 27 May 2024 13:23:53 +0200
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH v3] drm/nouveau/nvif: Avoid build error due to potential
 integer overflows
To: Guenter Roeck <linux@roeck-us.net>
Cc: Karol Herbst <kherbst@redhat.com>, Lyude Paul <lyude@redhat.com>,
 Daniel Vetter <daniel@ffwll.ch>, dri-devel@lists.freedesktop.org,
 nouveau@lists.freedesktop.org, linux-kernel@vger.kernel.org,
 Javier Martinez Canillas <javierm@redhat.com>,
 Jani Nikula <jani.nikula@intel.com>, Thomas Zimmermann
 <tzimmermann@suse.de>, Maxime Ripard <mripard@kernel.org>,
 Kees Cook <keescook@chromium.org>,
 Christophe JAILLET <christophe.jaillet@wanadoo.fr>,
 Joe Perches <joe@perches.com>, David Airlie <airlied@gmail.com>
References: <20240524134817.1369993-1-linux@roeck-us.net>
Content-Language: en-US
From: Danilo Krummrich <dakr@redhat.com>
Organization: RedHat
In-Reply-To: <20240524134817.1369993-1-linux@roeck-us.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit

On 5/24/24 15:48, Guenter Roeck wrote:
> Trying to build parisc:allmodconfig with gcc 12.x or later results
> in the following build error.
> 
> drivers/gpu/drm/nouveau/nvif/object.c: In function 'nvif_object_mthd':
> drivers/gpu/drm/nouveau/nvif/object.c:161:9: error:
> 	'memcpy' accessing 4294967264 or more bytes at offsets 0 and 32 overlaps 6442450881 bytes at offset -2147483617 [-Werror=restrict]
>    161 |         memcpy(data, args->mthd.data, size);
>        |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> drivers/gpu/drm/nouveau/nvif/object.c: In function 'nvif_object_ctor':
> drivers/gpu/drm/nouveau/nvif/object.c:298:17: error:
> 	'memcpy' accessing 4294967240 or more bytes at offsets 0 and 56 overlaps 6442450833 bytes at offset -2147483593 [-Werror=restrict]
>    298 |                 memcpy(data, args->new.data, size);
> 
> gcc assumes that 'sizeof(*args) + size' can overflow, which would result
> in the problem.
> 
> The problem is not new, only it is now no longer a warning but an error
> since W=1 has been enabled for the drm subsystem and since Werror is
> enabled for test builds.
> 
> Rearrange arithmetic and use check_add_overflow() for validating the
> allocation size to avoid the overflow. While at it, split assignments
> out of if conditions.
> 
> Fixes: a61ddb4393ad ("drm: enable (most) W=1 warnings by default across the subsystem")
> Cc: Javier Martinez Canillas <javierm@redhat.com>
> Cc: Jani Nikula <jani.nikula@intel.com>
> Cc: Thomas Zimmermann <tzimmermann@suse.de>
> Cc: Danilo Krummrich <dakr@redhat.com>
> Cc: Maxime Ripard <mripard@kernel.org>
> Cc: Kees Cook <keescook@chromium.org>
> Cc: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
> Cc: Joe Perches <joe@perches.com>
> Reviewed-by: Kees Cook <keescook@chromium.org>
> Signed-off-by: Guenter Roeck <linux@roeck-us.net>

Applied to drm-misc-fixes, thanks!

> ---
> v3: Split assignments from if conditions.
> v2: Use check_add_overflow() to calculate the allocation size and to check
>      for overflows.
> 
>   drivers/gpu/drm/nouveau/nvif/object.c | 24 ++++++++++++++++++------
>   1 file changed, 18 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/gpu/drm/nouveau/nvif/object.c b/drivers/gpu/drm/nouveau/nvif/object.c
> index 4d1aaee8fe15..1d19c87eaec1 100644
> --- a/drivers/gpu/drm/nouveau/nvif/object.c
> +++ b/drivers/gpu/drm/nouveau/nvif/object.c
> @@ -142,11 +142,16 @@ nvif_object_mthd(struct nvif_object *object, u32 mthd, void *data, u32 size)
>   		struct nvif_ioctl_v0 ioctl;
>   		struct nvif_ioctl_mthd_v0 mthd;
>   	} *args;
> +	u32 args_size;
>   	u8 stack[128];
>   	int ret;
>   
> -	if (sizeof(*args) + size > sizeof(stack)) {
> -		if (!(args = kmalloc(sizeof(*args) + size, GFP_KERNEL)))
> +	if (check_add_overflow(sizeof(*args), size, &args_size))
> +		return -ENOMEM;
> +
> +	if (args_size > sizeof(stack)) {
> +		args = kmalloc(args_size, GFP_KERNEL);
> +		if (!args)
>   			return -ENOMEM;
>   	} else {
>   		args = (void *)stack;
> @@ -157,7 +162,7 @@ nvif_object_mthd(struct nvif_object *object, u32 mthd, void *data, u32 size)
>   	args->mthd.method = mthd;
>   
>   	memcpy(args->mthd.data, data, size);
> -	ret = nvif_object_ioctl(object, args, sizeof(*args) + size, NULL);
> +	ret = nvif_object_ioctl(object, args, args_size, NULL);
>   	memcpy(data, args->mthd.data, size);
>   	if (args != (void *)stack)
>   		kfree(args);
> @@ -276,7 +281,15 @@ nvif_object_ctor(struct nvif_object *parent, const char *name, u32 handle,
>   	object->map.size = 0;
>   
>   	if (parent) {
> -		if (!(args = kmalloc(sizeof(*args) + size, GFP_KERNEL))) {
> +		u32 args_size;
> +
> +		if (check_add_overflow(sizeof(*args), size, &args_size)) {
> +			nvif_object_dtor(object);
> +			return -ENOMEM;
> +		}
> +
> +		args = kmalloc(args_size, GFP_KERNEL);
> +		if (!args) {
>   			nvif_object_dtor(object);
>   			return -ENOMEM;
>   		}
> @@ -293,8 +306,7 @@ nvif_object_ctor(struct nvif_object *parent, const char *name, u32 handle,
>   		args->new.oclass = oclass;
>   
>   		memcpy(args->new.data, data, size);
> -		ret = nvif_object_ioctl(parent, args, sizeof(*args) + size,
> -					&object->priv);
> +		ret = nvif_object_ioctl(parent, args, args_size, &object->priv);
>   		memcpy(data, args->new.data, size);
>   		kfree(args);
>   		if (ret == 0)