Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Wed, 8 May 2019 10:35:19 +0200
From:   Daniel Vetter <daniel@ffwll.ch>
To:     Jordan Crouse <jcrouse@codeaurora.org>
Cc:     freedreno@lists.freedesktop.org, linux-arm-msm@vger.kernel.org,
        Sean Paul <sean@poorly.run>, linux-kernel@vger.kernel.org,
        dri-devel@lists.freedesktop.org, David Airlie <airlied@linux.ie>,
        Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
        Maxime Ripard <maxime.ripard@bootlin.com>,
        Daniel Vetter <daniel@ffwll.ch>
Subject: Re: [PATCH] drm/atomic: Check that the config funcs exist
 drm_mode_alloc
Message-ID: <20190508083519.GS17751@phenom.ffwll.local>
Mail-Followup-To: Jordan Crouse <jcrouse@codeaurora.org>,
        freedreno@lists.freedesktop.org, linux-arm-msm@vger.kernel.org,
        Sean Paul <sean@poorly.run>, linux-kernel@vger.kernel.org,
        dri-devel@lists.freedesktop.org, David Airlie <airlied@linux.ie>,
        Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
        Maxime Ripard <maxime.ripard@bootlin.com>
References: <1557256451-24950-1-git-send-email-jcrouse@codeaurora.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1557256451-24950-1-git-send-email-jcrouse@codeaurora.org>
User-Agent: Mutt/1.10.1 (2018-07-13)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, May 07, 2019 at 01:14:11PM -0600, Jordan Crouse wrote:
> An error while initializing the msm driver ends up calling
> drm_atomic_helper_shutdown() without first initializing the funcs
> in mode_config. While I'm not 100% sure this isn't a ordering
> problem in msm adding a check to drm_mode_alloc seems like
> a nice and safe solution.
> 
> Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>

Hm yeah this looks a bit too much like ducttape. I think Noralf started
working on some ideas of devm-like automatic cleanup for drm stuff (we
cannot use devm, that has the wrong lifetimes, despite all the drivers
using it).

Simple fix would be to move up the assignment of config.funcs to be much
earlier in your driver load I guess.
-Daniel

> ---
> 
>  drivers/gpu/drm/drm_atomic.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
> index 5eb4013..1729428 100644
> --- a/drivers/gpu/drm/drm_atomic.c
> +++ b/drivers/gpu/drm/drm_atomic.c
> @@ -114,6 +114,9 @@ drm_atomic_state_alloc(struct drm_device *dev)
>  {
>  	struct drm_mode_config *config = &dev->mode_config;
>  
> +	if (!config->funcs)
> +		return NULL;
> +
>  	if (!config->funcs->atomic_state_alloc) {
>  		struct drm_atomic_state *state;
>  
> -- 
> 2.7.4
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch