linux-kernel - Re: [PATCH] drm/hisilicon: Added three new resolutions and changed the alignment to 128 Bytes

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

Date:   Mon, 30 Dec 2019 10:23:40 +0000
From:   Daniel Stone <daniel@...ishbar.org>
To:     Tian Tao <tiantao6@...ilicon.com>
Cc:     Chen Feng <puck.chen@...ilicon.com>,
        David Airlie <airlied@...ux.ie>,
        Daniel Vetter <daniel@...ll.ch>, tzimmermann@...e.de,
        Gerd Hoffmann <kraxel@...hat.com>,
        Alex Deucher <alexander.deucher@....com>, tglx@...utronix.de,
        dri-devel <dri-devel@...ts.freedesktop.org>,
        Xinliang Liu <xinliang.liu@...aro.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        linuxarm@...wei.com
Subject: Re: [PATCH] drm/hisilicon: Added three new resolutions and changed
 the alignment to 128 Bytes

Hi Tian,

On Sat, 28 Dec 2019 at 01:14, Tian Tao <tiantao6@...ilicon.com> wrote:
> @@ -118,11 +119,9 @@ static void hibmc_plane_atomic_update(struct drm_plane *plane,
>         writel(gpu_addr, priv->mmio + HIBMC_CRT_FB_ADDRESS);
>
>         reg = state->fb->width * (state->fb->format->cpp[0]);
> -       /* now line_pad is 16 */
> -       reg = PADDING(16, reg);
>
>         line_l = state->fb->width * state->fb->format->cpp[0];
> -       line_l = PADDING(16, line_l);
> +       line_l = PADDING(128, line_l);

The 'line length' here is the 'stride' field of the FB. Stride is set
by userspace when allocating the buffer, and the kernel must not
attempt to guess what userspace set.

You should use state->fb->strides[0] directly here, and in your
atomic_check() function, make sure that the framebuffer stride is
correctly aligned.

Please split this into a separate change. Your commit has three
changes in it, which should all be separate commits:
  * enforce 128-byte stride alignment (is this a hardware limit?)
  * get the BO from drm_fb rather than hibmc_fb (can hibmc_fb->obj
just be removed now?)
  * add new clock/resolution configurations

Cheers,
Daniel