[Nouveau] [PATCH] drm/nouveau: fix another off-by-one in nvbios_addr

[Karol Herbst]

On Wed, May 11, 2022 at 6:37 PM Timur Tabi <ttabi at nvidia.com> wrote:
>
> This check determines whether a given address is part of
> image 0 or image 1.  Image 1 starts at offset image0_size,
> so that address should be included.
>
> Fixes: 4d4e9907ff572 ("drm/nouveau/bios: guard against out-of-bounds accesses to image")
> Signed-off-by: Timur Tabi <ttabi at nvidia.com>
> ---
>  drivers/gpu/drm/nouveau/nvkm/subdev/bios/base.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/base.c
> index 64e423dddd9e..6c318e41bde0 100644
> --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/base.c
> +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/base.c
> @@ -33,7 +33,7 @@ nvbios_addr(struct nvkm_bios *bios, u32 *addr, u8 size)
>  {
>         u32 p = *addr;
>
> -       if (*addr > bios->image0_size && bios->imaged_addr) {
> +       if (*addr >= bios->image0_size && bios->imaged_addr) {
>                 *addr -= bios->image0_size;
>                 *addr += bios->imaged_addr;
>         }
> --
> 2.36.0
>

Reviewed-by: Karol Herbst <kherbst at redhat.com>

I don't like how this could fail for values spanning both images, but
at least this looks more correct.

I think in the future we should keep a copy without that gap on the
host side to make access to the vbios image cheaper. What do you say
Ben?