linux-kernel - Re: [PATCH] drm/panfrost: Fix leak when free gem object

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: <97149cf3-b3cb-4b90-a486-7eea5ee41b59@arm.com>
Date: Thu, 17 Jul 2025 11:23:24 +0100
From: Steven Price <steven.price@....com>
To: lihongtao <lihongtao@...inos.cn>,
 Boris Brezillon <boris.brezillon@...labora.com>
Cc: Rob Herring <robh@...nel.org>,
 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>,
 dri-devel@...ts.freedesktop.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] drm/panfrost: Fix leak when free gem object

On 10/07/2025 04:05, lihongtao wrote:
> obj->mappings.lock should be destroyed when free
> panfrost gem object in panfrost_gem_free_object.

mutex_destroy() doesn't actually release any resources. It is purely a
debugging feature (if CONFIG_DEBUG_MUTEXES is disabled then it is
compiled away completely).

So it's not a "leak" as such. But there is some value in using it as it
would (in debug builds) warn us if we attempt to destroy a locked mutex
or attempt to use a mutex after the destroy.

But if we're going to fix this I think we should be more complete. A
quick grep should we have 6 different mutexes in panfrost:

panfrost_device.c:      mutex_init(&pfdev->sched_lock);
panfrost_device.c:      mutex_init(&pfdev->debugfs.gems_lock);
panfrost_drv.c: mutex_init(&pfdev->shrinker_lock);
panfrost_gem.c: mutex_init(&obj->mappings.lock);
panfrost_gem.c: mutex_init(&obj->label.lock);
panfrost_perfcnt.c:     mutex_init(&perfcnt->lock);

But there's only one existing call to mutex_destroy():

panfrost_gem.c: mutex_destroy(&bo->label.lock);

It would be good to consider if the other mutexes should also be destroyed.

Thanks,
Steve

> 
> Signed-off-by: lihongtao <lihongtao@...inos.cn>
> ---
>  drivers/gpu/drm/panfrost/panfrost_gem.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/gpu/drm/panfrost/panfrost_gem.c b/drivers/gpu/drm/panfrost/panfrost_gem.c
> index 963f04ba2de6..00549f482eec 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_gem.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_gem.c
> @@ -49,6 +49,7 @@ static void panfrost_gem_free_object(struct drm_gem_object *obj)
>  		kvfree(bo->sgts);
>  	}
>  
> +	mutex_destroy(&bo->mappings.lock);
>  	drm_gem_shmem_free(&bo->base);
>  }
>