linux-kernel - Re: [Intel-gfx] [PATCH 16/26] drm/i915/gvt: use array

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 for Android: free password hash cracker in your pocket

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

Message-ID: <ZJlZ3421Whev/LkX@ashyti-mobl2.lan>
Date:   Mon, 26 Jun 2023 11:26:55 +0200
From:   Andi Shyti <andi.shyti@...ux.intel.com>
To:     Julia Lawall <Julia.Lawall@...ia.fr>
Cc:     Zhenyu Wang <zhenyuw@...ux.intel.com>, keescook@...omium.org,
        intel-gvt-dev@...ts.freedesktop.org,
        intel-gfx@...ts.freedesktop.org, kernel-janitors@...r.kernel.org,
        linux-kernel@...r.kernel.org, dri-devel@...ts.freedesktop.org,
        Daniel Vetter <daniel@...ll.ch>,
        Rodrigo Vivi <rodrigo.vivi@...el.com>,
        David Airlie <airlied@...il.com>
Subject: Re: [Intel-gfx] [PATCH 16/26] drm/i915/gvt: use array_size

Hi Julia,

On Fri, Jun 23, 2023 at 11:14:47PM +0200, Julia Lawall wrote:
> Use array_size to protect against multiplication overflows.
> 
> The changes were done using the following Coccinelle semantic patch:
> 
> // <smpl>
> @@
>     expression E1, E2;
>     constant C1, C2;
>     identifier alloc = {vmalloc,vzalloc};
> @@
>     
> (
>       alloc(C1 * C2,...)
> |
>       alloc(
> -           (E1) * (E2)
> +           array_size(E1, E2)
>       ,...)
> )
> // </smpl>
> 
> Signed-off-by: Julia Lawall <Julia.Lawall@...ia.fr>
> 
> ---
>  drivers/gpu/drm/i915/gvt/gtt.c |    6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c
> index 4ec85308379a..df52385ad436 100644
> --- a/drivers/gpu/drm/i915/gvt/gtt.c
> +++ b/drivers/gpu/drm/i915/gvt/gtt.c
> @@ -1969,14 +1969,16 @@ static struct intel_vgpu_mm *intel_vgpu_create_ggtt_mm(struct intel_vgpu *vgpu)
>  		return ERR_PTR(-ENOMEM);
>  	}
>  
> -	mm->ggtt_mm.host_ggtt_aperture = vzalloc((vgpu_aperture_sz(vgpu) >> PAGE_SHIFT) * sizeof(u64));
> +	mm->ggtt_mm.host_ggtt_aperture =
> +		vzalloc(array_size(vgpu_aperture_sz(vgpu) >> PAGE_SHIFT, sizeof(u64)));
>  	if (!mm->ggtt_mm.host_ggtt_aperture) {
>  		vfree(mm->ggtt_mm.virtual_ggtt);
>  		vgpu_free_mm(mm);
>  		return ERR_PTR(-ENOMEM);
>  	}
>  
> -	mm->ggtt_mm.host_ggtt_hidden = vzalloc((vgpu_hidden_sz(vgpu) >> PAGE_SHIFT) * sizeof(u64));
> +	mm->ggtt_mm.host_ggtt_hidden =
> +		vzalloc(array_size(vgpu_hidden_sz(vgpu) >> PAGE_SHIFT, sizeof(u64)));

thanks for this patch, but I see an issue here. array_size()
truncates the allocation to SIZE_MAX, and I'm OK with it.

The problem is that no error is notified and the user doesn't
know that a truncation has happened. So that if we save from an
overflow here, we might encur to an unwanted access later when we
would start using the array for the size we think is allocated.

kmalloc_array(), for example, returns NULL of there is a
multiplication overflow and I think that's a better behaviour,
although more drastic.

Andi

>  	if (!mm->ggtt_mm.host_ggtt_hidden) {
>  		vfree(mm->ggtt_mm.host_ggtt_aperture);
>  		vfree(mm->ggtt_mm.virtual_ggtt);