linux-kernel - Re: [RFC PATCH 1/4] media: Add Media Device Allocator API

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: <56F93589.4020102@osg.samsung.com>
Date:	Mon, 28 Mar 2016 07:45:45 -0600
From:	Shuah Khan <shuahkh@....samsung.com>
To:	Joe Perches <joe@...ches.com>, laurent.pinchart@...asonboard.com,
	mchehab@....samsung.com, perex@...ex.cz, tiwai@...e.com,
	hans.verkuil@...co.com, chehabrafael@...il.com,
	javier@....samsung.com, jh1009.sung@...sung.com
Cc:	linux-kernel@...r.kernel.org, linux-media@...r.kernel.org,
	alsa-devel@...a-project.org, Shuah Khan <shuahkh@....samsung.com>
Subject: Re: [RFC PATCH 1/4] media: Add Media Device Allocator API

On 03/26/2016 06:50 AM, Joe Perches wrote:
> On Fri, 2016-03-25 at 22:38 -0600, Shuah Khan wrote:
>> Add Media Device Allocator API to manage Media Device life time problems.
>> There are known problems with media device life time management. When media
>> device is released while an media ioctl is in progress, ioctls fail with
>> use-after-free errors and kernel hangs in some cases.
> 
> Seems reasonable, thanks.
> 
> trivial:
> 
>> diff --git a/drivers/media/media-dev-allocator.c b/drivers/media/media-dev-allocator.c
> []
>> +static struct media_device *__media_device_get(struct device *dev,
>> +					       bool alloc, bool kref)
>> +{
> []
>> +	pr_info("%s: mdev=%p\n", __func__, &mdi->mdev);
> 
> All of the pr_info uses here seem like debugging
> and should likely be pr_debug instead.

Correct. These are for debug and I plan to either remove them completely
or make them pr_debug().

>> +struct media_device *media_device_find(struct device *dev)
>> +{
>> +	pr_info("%s\n", __func__);
> 
> These seem like function tracing and maybe could/should
> use ftrace instead.
> +/* don't allocate - increment kref if one is found */
>> +struct media_device *media_device_get_ref(struct device *dev)
>> +{
>> +	pr_info("%s\n", __func__);
> 

Same here. This is also debug. However, you gave me an idea, this could be a
tracevent, if I find it useful for event tracing. It might be useful to be able
to track kref holds on this.

thanks,
-- Shuah