linux-kernel - Re: [PATCH 1/5] drm/solomon: Move calls to drm_gem_fb_end

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]

Message-ID: <874it1phx8.fsf@minerva.mail-host-address-is-not-set>
Date: Wed, 17 Sep 2025 11:06:11 +0200
From: Javier Martinez Canillas <javierm@...hat.com>
To: Iker Pedrosa <ikerpedrosam@...il.com>, Maarten Lankhorst
 <maarten.lankhorst@...ux.intel.com>, Maxime Ripard <mripard@...nel.org>,
 Thomas Zimmermann <tzimmermann@...e.de>, David Airlie <airlied@...il.com>,
 Simona Vetter <simona@...ll.ch>
Cc: dri-devel@...ts.freedesktop.org, linux-kernel@...r.kernel.org, Iker
 Pedrosa <ikerpedrosam@...il.com>
Subject: Re: [PATCH 1/5] drm/solomon: Move calls to drm_gem_fb_end_cpu*()

Iker Pedrosa <ikerpedrosam@...il.com> writes:

Hello Iker,

Thanks for your patch.

> Calls to drm_gem_fb_end_cpu*() should be between the calls to
> drm_dev*(), and not hidden inside some other function. This way the
> critical section code is visible at a glance, keeping it short and
> improving maintainability.
>
> Signed-off-by: Iker Pedrosa <ikerpedrosam@...il.com>
> ---
>  drivers/gpu/drm/solomon/ssd130x.c | 33 +++++++++++++++------------------
>  1 file changed, 15 insertions(+), 18 deletions(-)
>

[...]

> @@ -1232,6 +1214,9 @@ static void ssd130x_primary_plane_atomic_update(struct drm_plane *plane,
>  	if (!drm_dev_enter(drm, &idx))
>  		return;
>  
> +	if (drm_gem_fb_begin_cpu_access(fb, DMA_FROM_DEVICE))
> +		return;
> +

In this error path you should call drm_dev_exit(). The convention in the
kernel usually is to have a goto label for this, e.g.:

       if (drm_gem_fb_begin_cpu_access(fb, DMA_FROM_DEVICE))
              goto out_drm_dev_exit;

>  	drm_atomic_helper_damage_iter_init(&iter, old_plane_state, plane_state);
>  	drm_atomic_for_each_plane_damage(&iter, &damage) {
>  		dst_clip = plane_state->dst;
> @@ -1245,6 +1230,8 @@ static void ssd130x_primary_plane_atomic_update(struct drm_plane *plane,
>  				     &shadow_plane_state->fmtcnv_state);
>  	}
>  
> +	drm_gem_fb_end_cpu_access(fb, DMA_FROM_DEVICE);
> +

and then here before the call you could have the label.

out_drm_dev_exit:

>  	drm_dev_exit(idx);

Same comments for the other places where you are adding the
drm_gem_fb_end_cpu*() calls next to the drm_dev*() ones.

After the mentioned changes:

Reviewed-by: Javier Martinez Canillas <javierm@...hat.com>

-- 
Best regards,

Javier Martinez Canillas
Core Platforms
Red Hat