linux-kernel - Re: [PATCH] drm/fb-helper: add virtual screen size check to drm_fb_helper_check

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <CAMuHMdXRVHiBZj=MQ3jLy2AEnwrvMXt6KXbZXr10ZJMhJDAmsw@mail.gmail.com>
Date:   Thu, 11 Aug 2022 13:49:35 +0200
From:   Geert Uytterhoeven <geert@...ux-m68k.org>
To:     Andrey Strachuk <strochuk@...ras.ru>
Cc:     Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
        Maxime Ripard <mripard@...nel.org>,
        Thomas Zimmermann <tzimmermann@...e.de>,
        David Airlie <airlied@...ux.ie>,
        Daniel Vetter <daniel@...ll.ch>, Helge Deller <deller@....de>,
        DRI Development <dri-devel@...ts.freedesktop.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        ldv-project@...uxtesting.org
Subject: Re: [PATCH] drm/fb-helper: add virtual screen size check to drm_fb_helper_check_var()

Hi Andrey,

On Thu, Aug 11, 2022 at 12:23 PM Andrey Strachuk <strochuk@...ras.ru> wrote:
> Add virtual screen size check to drm_fb_helper_check_var() in
> order to validate userspace input.
>
> Found by Linux Verification Center (linuxtesting.org) with syzkaller.
>
> Signed-off-by: Andrey Strachuk <strochuk@...ras.ru>

Thanks for your patch!

> Fixes: 6c11df58fd1a ("fbmem: Check virtual screen sizes in fb_set_var()")

This Fixes tag is misleading: the bug has existed since basically
forever.

> --- a/drivers/gpu/drm/drm_fb_helper.c
> +++ b/drivers/gpu/drm/drm_fb_helper.c
> @@ -1355,6 +1355,16 @@ int drm_fb_helper_check_var(struct fb_var_screeninfo *var,
>             (drm_format_info_block_height(fb->format, 0) > 1))
>                 return -EINVAL;
>
> +       /* verify that virtual resolution >= physical resolution */
> +       if (var->xres_virtual < var->xres ||
> +           var->yres_virtual < var->yres) {
> +               drm_dbg_kms(dev, "requested virtual screen size that is "
> +                               "below the physical size(%ux%u vs. %ux%u)\n",
> +                       var->xres_virtual, var->yres_virtual,
> +                       var->xres, var->yres);
> +               return -EINVAL;

According to the fbdev rounding rules, invalid values should be rounded
up, if possible.  Hence it is better to not return an error, but round up:

    if (var->xres_virtual < var->xres)
            var->xres_virtual = var->xres;
    if (var->yres_virtual < var->yres)
            var->yres_virtual = var->yres;

> +       }
> +
>         /*
>          * Changes struct fb_var_screeninfo are currently not pushed back
>          * to KMS, hence fail if different settings are requested.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@...ux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds